Interface Controlled Organic Thin Films

springer proceedings in physics 129

springer proceedings in physics 113 Theoretical and Numerical Unsaturated Soil Mechanics Editor: T. Schanz 114 Advances in Medical Engineering Editor: T.M. Buzug 115 X-Ray Lasers 2006 Proceedings of the 10th International Conference, August 20–25, 2006, Berlin, Germany Editors: P.V. Nickles, K.A. Janulewicz 116 Lasers in the Conservation of Artworks LACONA VI Proceedings, Vienna, Austria, Sept. 21–25, 2005 Editors: J. Nimmrichter, W. Kautek, M. Schreiner 117 Advances in Turbulence XI Proceedings of the 11th EUROMECH European Turbulence Conference, June 25–28, 2007, Porto, Portugal Editors: J.M.L.M. Palma and A. Silva Lopes 118 The Standard Model and Beyond Proceedings of the 2nd International Summer School in High Energy Physics, M¯gla, 25–30 September 2006 Editors: M. Serin, T. Aliev, N.K. Pak 119 Narrow Gap Semiconductors 2007 Proceedings of the 13th International Conference, 8–12 July, 2007, Guildford, UK Editors: B. Murdin, S. Clowes 120 Microscopy of Semiconducting Materials 2007 Proceedings of the 15th Conference, 2–5 April 2007, Cambridge, UK Editors: A.G. Cullis, P.A. Midgley

122 Advances in Nanoscale Magnetism Proceedings of the International Conference on Nanoscale Magnetism ICNM-2007, June 25–29, Istanbul, Turkey Editors: B. Aktas, F. Mikailov 123 Computer Simulation Studies in Condensed-Matter Physics XIX Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 124 EKC2008 Proceedings of the EU-Korea Conference on Science and Technology Editor: S.-D. Yoo 125 Computer Simulation Studies in Condensed-Matter Physics XX Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 126 Vibration Problems ICOVP 2007 Editors: E. Inan, D. Sengupta, M.M. Banerjee, B. Mukhopadhyay, and H. Demiray 127 Physics and Engineering of New Materials Editors: D.T. Cat, A. Pucci, and K.R. Wandelt 128 Ultrasonic Wave Propagation in Non Homogeneous Media Editors: A. Leger, M. Deschamps 129 Interface Controlled Organic Thin Films Editors: H.-G. Rubahn, H. Sitter, G. Horowitz, K. Al-Shamery

121 Time Domain Methods in Electrodynamics A Tribute to Wolfgang J. R. Hoefer Editors: P. Russer, U. Siart

Volumes 90–112 are listed at the end of the book.

¨ Horst-Gunter Rubahn Helmut Sitter Giles Horowitz Katharina Al-Shamery Editors

Interface Controlled Organic Thin Films With 122 Figures

123

Prof. Dr. Horst-Günter Rubahn

Prof. Dr. Helmut Sitter

University of Southern Denmark NanoSYD Mads Clausen Institute Alsion 2, 6400 Sønderborg Denmark [email protected]

Universität Linz Institut für Experimentalphysik Abteilung Festkörperphysik 4040 Linz Austria [email protected]

Prof. Dr. Giles Horowitz

Prof. Dr. Katharina Al-Shamery

Université Paris VII Institut Topologie et de Dynamique Systèmes (ITODYS) 1 rue Guy-de-la-Brosse 75005 Paris France [email protected]

Universität Oldenburg Institut für Reine und Angewandte Chemie 26111 Oldenburg Germany [email protected]

ISSN 0930-8989 ISBN 978-3-540-95929-8 e-ISBN 978-3-540-95930-4 DOI 10.1007/978-3-540-95930-4 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: “PCN applied for” © Springer-Verlag Berlin Heidelberg 2009 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: SPi Publisher Services Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface Organic electronics is a scientific and technological field that has witnessed an enormous world-wide effort both in basic scientific research as well as in industrial development within the last decades. It is becoming increasingly clear that, if devices based on organic materials are ever going to have a significant relevance beyond being a cheap replacement for inorganic semiconductors, there will be a need to understand interface formation, film growth and functionality. A control of these aspects will allow the realisation of totally new device concepts exploiting the vast flexibility inherent in organic chemistry. The field of devicerelevant “semiconducting” organic materials has many parallels to that of inorganic semiconductors. However, the versatility of organic molecules comes at the cost of higher complexity of the materials. This rules out a 1:1 transfer of concepts established within inorganic semiconductor research to the world of organics, and makes work on organic semiconductors particularly challenging. On a world-wide scale, investigations of organic thin films focus on three main areas with different aims and with a fruitful mixture of applied and basic research: (1) the development and production of devices, (2) thin film characterization and more recently, after recognizing the importance of molecular level control (3) surface and interface science. Linking these branches together creates new synergies and has led and leads to a significant advance in the field of organic semiconductors. Eventually it will result in the development of the necessary tools for tuning device properties on a nanoscopic level. In the last 10 to 15 years a large amount of investigations of devices have been performed with a big range of active organic materials. This work has mapped out the classes of materials that proof useful for single molecule, oligomeric/molecular films and plastic electronics. In this symposium we focused on oligomeric/molecular films, because the control of molecular structures and interfaces provides unprecedently highly defined systems. This in turn allows one to study basic physics and at the same time enables one to find the important parameters necessary to improve organic devices. The E-MRS symposium conceived to bring together the leading groups, which work in the field of growth and characterisation of organic films and devices and focus them on the fabrication and characterisation of highly ordered functional organic films. The wide range of expertise of the contributing groups allowed the combination of different methodologies and aspects of physics, chemistry, and materials science for the design and understanding of well-defined organic structures. In total we received 148 contributions to the symposium in the form of invited talks, oral presentations and posters. Out of them the reviewers selected a representative amount of papers to be published in the proceedings. The main topics discussed at the symposium are reflected in the headlines of the chapters in the proceedings. Introductory review papers based on invited talks given at the symposium are followed by contributed papers. The highlights of the oral and poster presentations contributing to the same topic are summarized in the same chapter. The editors would like to thank the sponsors of the E-MRS symposium, especially the ‘Fonds der Chemischen Industrie’.

vi

Preface

In technical finishing the book we would like to thank Ms. Zora Milde for her extraordinary help in mastering the handling of all the electronic documents.

Sønderborg, Linz, Paris, Oldenburg January 2009

H.-G. Rubahn H. Sitter G. Horowitz K. Al-Shamery

Contents Preface .................................................................................................................... V A Thin Film Growth............................................................................................ 1 1

Toward an Ab-initio Description of Organic Thin Film Growth ............... 3 P. Puschnig, D. Nabok, and C. Ambrosch-Draxl

2

Organic Nano Fibres from PPTPP .............................................................. 11 F. Balzer, M. Schiek, A. Lützeu, and H.G. Rubahn

3

α-Sexithiophene Films Grown on Cu(110)–(2x1)O: From Monolayer to Multilayers ................................................................................................. 19 M. Oehzelt,, S. Berkebile, G. Koller, T. Haber, M. Koini, O. Werzer, R. Resel, and M.G. Ramsey

4

Para-Sexiphenyl Layers Grown on Light Sensitive Polymer Substrates ....................................................................................................... 23 G. Hernandez-Sosa, C. Simbrunner, T. Höfler, A. Moser, O. Werzer, B. Kunert, G. Trimmel, W. Kern, R. Resel and H. Sitter

5

Thermal Desorption of Organic Molecules................................................. 29 A. Winkler

6

Crystalline Stages of Rubrene Films Probed by Raman Spectroscopy................................................................................................... 37 B.A. Paez, Sh. Abd-Al-Baqi, G.H. Sosa, A. Andreev, C. Winder, F. Padinger, C. Simbrunner, and H. Sitter

7

Rubrene Thin Film Characteristics on Mica .............................................. 43 Sh.M. Abd Al-Baqi, G. Henandez-Sosa, H. Sitter, B. Th. Singh, Ph. Stadler, N.S. Sariciftci

8

Structural Properties of Rubrene Thin Films Grown on Mica ........................................................................................................ ...49 T. Djuric, H.-G. Flesch, M. Koini, Sh.M. Abd Al-Baqi, H. Sitter,and R. Resel

vii

viii

9

Contents

Rubrene on Mica: From The Early Growth Stage To Late Crystallization................................................................................................ 55 G. Hlawacek, S. Abd-al Baqi, X. Ming He, H. Sitter, and C. Teichert

10 β-Sheeted Amyloid Fibril Based Structures for Hybrid Nanoobjects on Solid Surfaces......................................................................................... ..61 V. Bukauskas, V. Strazdienė, A. Šetkus, S. Bružytė, V. Časaitė,and R. Meškys 11 Characteristics of Vacuum Deposited Sucrose Thin Films .................... ..67 F. Ungureanu, D. Predoi, R.V. Ghita, R.A. Vatasescu-Balcan,and M. Costache 12 Electropolymerization of Polypyrrole Films in Aqueous Solution with Side-Coupler Agent to Hydrophobic Groups.................................. ..73 H.M. Alfaro-López, J.R. Aguilar-Hernandez, A. Garcia-Borquez, M.A. Hernandez-Perez,and G.S. Contreras-Puente. 13 Surface Modification of Polymer Powders by a Far Cold Remote Nitrogen Plasma in Fluidized Bed............................................................. ..79 L. Aiche, H. Vergnes, B. Despax, B. Caussat,and H. Caquineau 14 Features of Polytetrafluoroethylene Coating Growth on Activated Surfaces from Gas Phase .......................................................................... ..85 A.A. Rogachev, S. Tamulevičius, A.V. Rogachev, I. Prosycevas,and M. Andrulevičius 15 Modification of Amorphous Carbon Film Surfaces by Thermal Grafting of Alkene Molecules.................................................................... ..91 H. Sabbah, A. Zebda, S. Ababou-Girard, B. Fabre, S. Députier, A. Perrin, M. Guilloux-Viry,and F. Solal, C. Godet 16 DNA-Controlled Assemblage of Ag Nanoparticles on Solid Surfaces .. ..95 V. Bukauskas, A. Šetkus, I. Šimkienė, J. Sabataitytė, A. Kindurys, and A. Rėza, J. Babonas

Contents

ix

17 Characterization of Self Assembled Monolayer Formation of 11 - Mercaptoundecanoic Acid on Gold Surfaces .............................. 101 J. Stettner, P. Frank, T. Griesser, G. Trimmel, R. Schennach, R. Resel, and A. Winkler 18 SAMs of 11-MUA Grown on Polycrystalline Au-foils by Physical Vapor Deposition in UHV...................................................................................... 107 P. Frank, F. Nussbacher, J. Stettner,and A. Winkler 19 Photoreactive Self Assembled Monolayers for Tuning the Surface Polarity ........................................................................................................ 113 T. Griesser, A. Track, G. Koller, M. Ramsey W. Kern, and G. Trimmel B Traps and Defects ....................................................................................... 119 20 Spectroscopy of Defects in Epitaxially Grown Para-sexiphenyl Nanostructures............................................................................................ 121 A. Kadashchuk, S. Schols, Yu. Skryshevski, I. Beynik, C. Teichert , G. Hernandez-Sosa, H. Sitter, A. Andreev, P. Frank, and A. Winkler 21 Magnetoresistance in Poly (3-hexyl thiophene) Based Diodes and Bulk Heterojunction Solar Cells........................................................ 127 S. Majumdar, H. S. Majumdar, H. Aarnio, R. Laiho, and R. Österbacka 22 Evolution of the Bipolaron Structure in Oligo-diacetylene Films: A Semiempirical Study .............................................................................. 133 M. Ottonelli, G. Musso, and G. Dellepiane C Energy Level Alignment and Charge Transfer ....................................... 139 23 Molecular Orientation Dependence of the Ionization Energy of Pentacene in Thin Films ........................................................................ 141 G. Heimel and N. Koch 24 Charge transfer and Polarization Screening at Organic/Metal Interfaces: Single Crystalline versus Polycrystalline Gold .................... 147 H. Peisert, D. Kolacyak, A. Petershans,and T. Chassé

x

Contents

25 Sensing Infrared Light with an Organic/Inorganic Hetero-junction.... 153 Gebhard J. Matt, Thomas Fromherz , Guillaume Goncalves, Christoph Lungenschmied, Dieter Meissner, and Serdar N. Sariciftci D Advanced Characterization Methods ....................................................... 159 26 Ultrafast Confocal Microscope for Functional Imaging of Organic Thin Films ................................................................................................... 161 Dario Polli, Michele Celebrano, Jenny Clark, Giulia Grancini, Tersilla Virgili, Guglielmo Lanzani, and Giulio Cerullo 27 Growth and Desorption Kinetics of Sexiphenyl Needles: An In-situ AFM/PEEM Study ..................................................................................... 167 Alexander J. Fleming, Svetlozar Surnev, Falko P. Netzer, and Michael G. Ramsey E

Organic Devices .......................................................................................... 171

28 Temperature Dependence of the Charge Transport in a C60 based Organic Field Effect Transistor ................................................................ 173 Mujeeb Ullah, Th.B. Singh, G.J. Matt, C. Simbruner, G. Hernandz-Sosa, S.N. Sariciftci,and H. Sitter 29 The Influence of Chain Orientation in the Electric Behaviour of Polymer Diodes........................................................................................................... 179 Marta Ramos and Helder Barbosa 30 Interface Modification of Pentacene Ofet Gate Dielectrics .................... 185 Ján Jakabovič, Jaroslav Kováč, Rudolf Srnánek, Jaroslav Kováč jr., Michal Sokolský, Július Cirák, Daniel Haško, Roland Resel, and Egbert Zojer 31 Negative Differential Resistance in C60 Diodes ........................................ 189 Philipp Stadler, Anita Fuchsbauer, Günther Hesser, Thomas Fromherz, Gebhard J. Matt, Helmut Neugebauer, and Serdar N. Sariciftci 32 Performance and Transport Properties of Phthalocyanine: Fullerene Organic Solar Cells ................................................................... 195 M. Rusu, J. Gasiorowski, S. Wiesner, D. Keiper, N. Meyer, M. Heuken, K. Fostiropoulos, and M.Ch. Lux-Steiner1

Contents

xi

33 Organic Transistors Based on Molecular and Polymeric Dielectric Materials...................................................................................................... 199 A. Facchetti, S. DiBenedetto, C. Kim,and T.J. Marks 34 Morphology of the Metal-Organic Semiconductor Contacts: the Role of Substrate Surface Treatment ................................................ 205 A.Petrović, E. Pavlica, and G. Bratina 35 Molecular Interactions Between Alcohols and Metal Phthalocyanine Thin Films for Optical Gas Sensor Applications..................................... 211 S. Uttiya, S. Kladsomboon, O. Chamlek, W. Suwannet, T. Osotchan, T. Kerdcharoen, M. Brinkmann, and S. Pratontep 36 Organic Thin-Film Transistors with Enhanced Sensing Capabilities.................................................................................................. 217 M. Daniela Angione, F. Marinelli, A. Dell’Aquila, A. Luzio , B. Pignataro and L. Torsi 37 Photoelectric Properties of Microrelief Organic/Inorganic Semiconductor Heterojunctions................................................................ 225 N.L. Dmitruk, O.Yu. Borkovskaya, D.O. Naumenko, I.B. Mamontova, N.V. Kotova, O.S. Lytvyn, and Ya.I. Vertsimakha List of Contributors........................................................................................... 229

Toward an Ab-initio Description of Organic Thin Film Growth Peter Puschnig, Dmitrii Nabok and Claudia Ambrosch-Draxl Chair of Atomistic Modelling and Design of Materials, Montanuniversität Leoben, Franz-Josef-Straße 18, A-8700 Leoben, Austria E-mail: [email protected] Abstract. We present an overview of recent ab initio calculations towards the modeling of organic thin film growth. First, we address the intermolecular bonding properties of the oligoacene, oligophenylene and oligothiophene series by density functional theory. By including non-local correlations to account for the van der Waals interactions we achieve excellent agreement of the cohesive energies with available experimental data and obtain surface energies for various low-index planes, thereby emphasizing the importance of dispersive interactions. Second, we review the findings for the interface energy of the model organic/metal interface, thiophene/Cu(110), using the same methodology. Finally, we show how a combination of ab inito results with an empirical force field approach leads to diffusion barriers relevant for organic film growth.

1.

Introduction

In the past years considerable experimental efforts have been placed toward controlling thin film morphologies organic semiconductors by tuning thin film growth conditions [1–3]. A defect-free layer-by-layer growth with a desired orientation of the molecules on a substrate is the key for an optimized device performance. The microscopic quantities which are determining the growth morphologies are surface and interface energies, as well as kinetic parameters, e.g. intra- and interlayer diffusion barriers. However, these numbers are difficult to access from experiments, and hence a theoretical approach for the energetics governing intermolecular as well as molecule/substrate interactions proves particularly useful. Owing to the advances in computer technology first-principles density functional calculations for complex molecular crystals have come into reach. Moreover, recent advances in density functional theory (DFT) have allowed for the incorporation of non-local dispersion forces into the DFT framework [4]. This van der Waals density functional (vdWDF) functional has proven to yield reliable binding energies for a wide class of typical van-der-Waals (vdW) bound systems ranging from noble gas dimers [4], and graphite [5], to the adsorption of benzene and naphtalene on graphite [6], polyethylene crystal structure [7], and the adsorption of thiophene on a metallic surface [8]. Thereby, the viability of the ab initio DFT approach for a number of quantities relevant for organic film growth has been demonstrated. In this article, we show three examples of how state-of-the-art first-principles calculations can contribute to the understanding of organic thin film growth. First, we discuss the cohesive and surface energies of three important types of π-conjugated H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_1, © Springer-Verlag Berlin Heidelberg 2009

3

4

P. Puschnig, D. Nabok and C. Ambrosch-Draxl

molecules, i.e., oligoacenes (nA), oligophenylenes (nP), and oligothiophenes (nT) where n denotes to the number of monomer units [9]. We make use of the vdWDF [4] and obtain cohesive energies in excellent agreement with experimental data and provide a first-principles prediction of surface energies. The latter allows us to estimate the equilibrium crystal shapes for all studied compounds. Second, we show results for the interface energy of a model organic/metal junction, thiophene/Cu(110) [8], and demonstrate its implications for resulting growth modes. Finally, we review recent findings for the step-edge barrier observed in growth of para-sexiphenyl [10].

2.

Computational Details

Density Functional Calculations. The total-energy and force calculations are performed within the framework of DFT as implemented in the plane-wave package PWSCF [11] using ultrasoft pseudo-potentials [12] with a plane-wave energy cutoff of 40 Ry. Bulk energies are computed by adopting the experimentally known space groups and lattice constants and relaxing the internal atomic positions. Isolated molecules are treated by supercells, while surface energies are calculated by the repeated-slab approach. Only one molecular layer turned out to be sufficient to converge surface energies to within 10 meV. Vacuum distances of 10 Å were found to be adequate to prevent interaction between translational images for isolated molecules as well as slab geometries. We use three different approximations for the exchange-correlation energy. Apart from the standard functionals like the local density approximation (LDA) [13] and the generalized gradient approximation (GGA) [14] we also employ the van der Waals density functional (vdWDF) in a non-selfconsistent approach [4]. The molecular cohesive energy Ecoh is defined as the energy reduction upon forming a crystal from isolated molecules, i.e., Ecoh = Emol – Ebulk/2, where Ebulk and Emol denote the total energies of the bulk and the isolated molecule, respectively. The factor 2 takes into account the number of molecules in the unit cell. By this definition the cohesive energy is positive for any stable crystal. The surface energy is defined as the energy required for cleaving a surface from a bulk material. The surface energy can be specified either per surface unit cell as Esurf = ½(Eslab – Ebulk), or per surface area A as γ = ½(Eslab – Ebulk)/A. Here, Eslab denotes the total energy of the slab configuration, and the factor ½ stands for the fact that the slab contains two surfaces.

3.

Results

Cohesive Energies. The bulk phases of the nA, nP, and nT series are characterized by the so-called herringbone packing where two inequivalent molecules in the ab plane form layers perpendicular to the crystallographic c direction. At room temperature and ambient pressure the space group of the considered molecular crystals is monoclinic, except for tetracene (4A) and pentacene (5A) which crystallize in a triclinic space group.

Toward an Ab-initio Description of Organic Thin Film Growth

5

Cohesive energies for the oligoacene series are presented in Figure 1 where DFT results obtained within the LDA, the GGA and the vdWDF are compared to measured data. For the whole oligoacene series we find excellent agreement between experimental and vdWDF values. The LDA results systematically underestimate the measurement by almost 25% while GGA yields practically zero cohesive energy. This demonstrates the inherent problem of local (LDA) or semilocal functionals (GGA) when applied to weakly bound systems, namely, that these standard approximations do not capture the essential physics of van der Waals interactions requiring a truly non-local functional. Even though LDA reproduces the experimental trend in a reasonable manner [15] it should not be used for predicting intermolecular binding properties since it relies on the wrong physical picture. Only through its general overbinding effect does it mimic a compensation for missing dispersive interactions. It is noteworthy that the cohesive energy is an almost linear function of molecular length which appears reasonable owing to the same type of crystalline packing for the whole oligomer series. The situation is similar for the oligophenylenes, but somewhat more complicated. Here we have to consider the non-planar nature of the isolated nP molecules, which exhibit twisted interring bonds in the gas phase. Allowing for such a conformation of the isolated molecule leads to a smaller Ecoh compared to the assumption of planar molecules. For instance, the effect on the cohesive energy of biphenyl is found to be 0.1 eV, where the torsion angle corresponding to the energy minimum is about 37°, and we obtain cohesive energies of 0.98, 1.44, 1.90, and 2.82 eV for 2P, 3P, 4P, and 6P, respectively. Finally, also the oligothiophenes show a similar trend with respect to oligomer length as the nA and the nP series. Since there is no experimental data available for the thiophene oligomers, our vdW-DF data – 0.95, 1.75, and 2.78 eV for 2T, 4T, and 6T, respectively – can be viewed as an accurate theoretical prediction.

Fig. 1. Cohesive energies of the oligoacene series from napthtalene (2A) to pentacene (5A). Theoretical values obtained within the van-der-Waals density functional (vdW-DF), the local densiy approximation (LDA), and the generalized gradient approximation (GGA) are compared to experimental data.

6

P. Puschnig, D. Nabok and C. Ambrosch-Draxl

Relating the cohesive energy to the number of carbon atoms per molecule one finds very similar results for the acenes (85, 79, 79, and 76 meV/C atom) and phenylenes (82, 80, 79, 79 meV/C atom), while the presence of sulfur atoms in the thiophenes increases the cohesive energies by roughly 10%. Surface Energies. Concerning organic thin film morphologies, the anisotropy in the surface energies is an important quantity controlling the orientation and shape of crystals during the growth. The (100), (010), (001), and (110) surface energies for the complete nA, nP, and nT series are given in Figure 2. A common feature for all compounds is that the (001) plane exhibits the lowest surface energy. This can be understood in terms of the crystal packing in that direction which is governed by rather weak H—H interlayer interactions. Hence, on substrates with comparably small substrate/molecule interactions thin films are expected to wet the substrate and to be preferentially (001) oriented since in that way the total surface free energy is minimized.

Fig. 2. Surface energies within the vdWDF for the oligoacene (nA), oligophenylene (nP), and the oligothiophene (nT) series. Values for the (100), (010), (001), and (110) planes are given and the corresponding surface terminations are depicted for tetracene (4A).

This fact is also illustrated by the equilibrium crystal shapes (ECS) as exemplified in Figure 3 for 3A where the (001) crystal faces have the largest areas. These are based on Wulff’s construction using the surface energies as summarized in Figure 2. Compared to experiment we find excellent agreement of our equilibrium crystal shape with a recent investigation on the growth of anthracene on graphite [16]. In particular the (001) orientation of the crystal parallel to the substrate and the appearance of approximately hexagonally shaped facets of (100) and (110) planes and a small (010) facet is in line with our calculated ECS.

Toward an Ab-initio Description of Organic Thin Film Growth

7

Fig. 3. Calculated equilibrium crystal shape of anthracene (filled octagon) compared to a photograph of the epitaxial growth of an anthracene single crystal on a graphite (0001) substrate taken from Ref. [16]. Interface Energies. In the previous section we have learned that characteristic surface energies of organic molecular crystals are in the range of 80–160 mJ/m2. For growth on inorganic substrates these numbers for the adsorbate surface energy γa are to be compared to typical surface energies of the inorganic substrate γs which are typically an order of magnitude larger. This is of course due to the fact that in inorganic materials strong covalent bonds have to be broken at the surface while in organic molecular crystals only comparably weak van der Waals forces have to be overcome to create a surface from the bulk. Close to thermodynamic equilibrium the magnitude of the substrate surface energy γs, the adsorbate surface energy γa, and the interface energy γi determine the resulting growth mode by the requirement of minimizing the total free energy. Therefore, it is important to calculate the organic/inorganic interface energy by employing a reliable and accurate ab-initio approach. As a model system we have studied the interaction between a thiophene molecule and the Cu(110) surface [8]. By employing vdWDF as previously for the investigation of the surface energies we have found van der Waals interactions to play a crucial role also for this organic/metal junction. This is illustrated in Figure 4 where we compare DFT results obtained within the LDA and GGA to the vdWDF results. We find an adsorption energy of Ea = 0.5 eV which is almost twice as large as the corresponding GGA value, and more than two times smaller than the adsorption energy predicted by LDA. We should emphasize the fact that our vdWDF value is in good agreement with the experimental Ea reported for thiophene on Cu(100) [17]. Using the definition, γi = γs + γa – Ea/A, for the interface energy as sum of the substrate and adsorbate surface energies minus the adsorption energy per area A we arrive at a value of γi = 1.55 J/m2 for the thiophene/Cu(110) interface. For this finding we have taken into account the surface energy of Cu(110) of γs = 1.70 mJ/m2 and approximated the surface energy of thiophene γa by the (100) value of bithiophene which is 0.15 mJ/m2 according to Figure 2.

8

P. Puschnig, D. Nabok and C. Ambrosch-Draxl

Fig. 4. Left: Adsorption energy of a thiophene ring on a Cu(110) surface calculated within the LDA, the GGA, and the van der Waals density functional [8]. Right: Schematic representation of the thiophene/Cu(110) adsorption.

In inorganic film growth the difference Δγ = γa + γi – γs is known to be a good indicator for the type of growth mode. A negative value for Δγ leads to layerby-layer (Frank van der Merve) growth since it is energetically favorable to cover the substrate completely by the adsorbate, while a positive value points towards three-dimensional island (Volmer-Weber) growth [18]. Evaluating Δγ for the thiophene/Cu(110) model system we arrive at a value close to 0. This indicates the intermediate case referred to as Stranski-Krastanov growth mode in which layerby-layer growth is succeeded by three-dimensional island growth after a critical adsorbate thickness. While in inorganic film growth the critical layer thickness depends on the strain built up in the deposited film, the situation in organic growth is more complicated due to the anisotropy and the internal degrees of freedom of the organic building blocks. For instance, wetting layers having various molecular conformation or orientation may be formed compared to the molecular orientation at later growth stages. Also, taking into account values for Δγ ≅ 0 as observed for the thiophene/Cu(110) model system we can expect a diversity of growth morphologies as is indeed observed in organic film growth. Kinetic Barriers. The discussion above assumes that the energetically lowest configuration can indeed be reached during the thin film growth process. However, kinetic barriers often limit the mobility of the deposited molecules and give rise to additional phenomena. In this section we exemplify the consequence of a step edge-barrier on the resulting growth morphology for the growth of 6P on a modified mica surface. According to Figure 2, the crystal plane exhibiting the lowest surface energy in 6P is the (001) plane resulting in γa(001) = 0.11 eV. Since the interaction between a (001) terminated 6P layer and the mica substrate is expected to be small one can assume the interface energy for that system to be smaller than the surface energy of the mica substrates. This would result in a layer-by-layer growth of upright standing 6P molecules. Instead one observes

Toward an Ab-initio Description of Organic Thin Film Growth

9

the formation of terraced mounds which can be explained by the existence of an additional step-edge barrier (Ehrlich-Schwoebel barrier). An analysis of these growth mounds as a function of coverage by atomic force microscopy revealed a sizeable Ehrlich-Schwoebel barrier (ESB) of 0.67 eV [10]. By combining firstprinciples density functional calculations with an empirical force field method we were able to calculate the total activation barrier to be 0.63 eV. By taking into account the computed on-terrace diffusion barrier of only 0.02 eV, the ESB results in 0.61 eV which is in excellent agreement with the experimental value [10]. Moreover, the calculation of the transition state provided further insight into the nature of the step-edge diffusion barrier, by showing that the 6P molecule diffuses toward the [100] edge with its long axis perpendicular to the edge and gradually slides down the (100) plane by bending over the edge between (001) and (100) facets. This is illustrated in Figure 5 which depicts four snap shots of the diffusion path for the step-edge crossing.

Fig. 5. Four snapshots of the diffusion path for step-edge crossing of sexiphenyl on a (001) 6P surface. The energy of each configuration is indicated, where the second structure corresponds to the transition state with an activation barrier of 0.63 eV.

4.

Conclusions

In summary, we have presented an ab-initio study of bonding properties in molecular crystals by consistently taking into account non-local van der Waals interactions. We reliably produce correct equilibrium layer distances to within 0.2 Å. This new level of precision enables us to obtain surface energies with an estimated accuracy of 5% allowing for accurate predictions of equilibrium crystal shapes and thin film morphologies where subtle energy differences play an important role. Moreover, our results open the perspective towards future investigations aiming at kinetic parameters for organic thin film growth. Acknowledgements. The work was supported by the Austrian Science Fund, Project No. S9714 within the National Science Network “Interface Controlled and Functionalised Organic Films”.

10

P. Puschnig, D. Nabok and C. Ambrosch-Draxl

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

F.-J. M. zu Heringdorf, M. C. Reuter, and R. M. Tromp, Nature 412, 51, 2001. M. A. Loi, E. da Como, F. Dinelli, M. Murgia, R. Zamboni, F. Biscarini, and M. Muccini, Nat. Mat. 4, 81, 2005. G. Hlawacek, Q. Shen, C. Teichert, R. Resel, and D. M. Smilgies, Surf. Sci. 601, 2584, 2007. M. Dion, H. Rydberg, E. Schröder, D. C. Langreth, and B. I. Lundqvist, Phys. Rev. Lett. 92, 246401, 2004. H. Rydberg, M. Dion, N. Jacobson, E. Schröder, P. Hyldgaard, S. I. Simak, D. C. Langreth, and B. I. Lundqvist, Phys. Rev. Lett. 91, 126402, 2003. S. D. Chakarova-Käck, E. Schröder, B. I. Lundqvist, and D. C. Langreth, Phys. Rev. Lett. 96, 146107, 2006. J. Kleis, B. I. Lundqvist, D. C. Langreth, and E. Schröder, Phys. Rev. B 76, 100201(R), 2007. P. Sony, P. Puschnig, D. Nabok, and C. Ambrosch-Draxl, Phys. Rev. Lett. 99, 176401, 2007. D. Nabok, P. Puschnig, and C. Ambrosch-Draxl, Phys. Rev. B 77, 245316, 2008. G. Hlawacek, P. Puschnig, P. Frank, A. Winkler, C. Ambrosch-Draxl, and C. Teichert, Science 321, 108, 2008. S. Baroni, A. D. Corso, S. de Gironcoli, P. Giannozzi, C. Cavazzoni, G. Ballabio, S. Scandolo, G. Chiarotti, P. Focher, and A. Pasquarello, http://www.pwscf.org/, 2007. D. Vanderbilt, Phys. Rev. B 41, 7892, 1990. J. P. Perdew and A. Zunger, Phys. Rev. B 23, 5048, 1981. J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865, 1996). J. E. Northrup, M. L. Tiago, and S. G. Louie, Phys. Rev. B 66, 121404(R), 2002. S. Jo, H. Yoshikawa, A. Fujii, and M. Takenaga, Surf. Sci. 592, 37, 2005. B. A. Sexton, Surf. Sci. 163, 99, 1985. A. Groß, Theoretical Surface Science – A Microscopic Perspective, Springer, Berlin, 2003.

Organic Nanofibers from PPTPP Frank Balzer1, Manuela Schiek1, Arne Lützeu2 and Horst-Günter Rubahn1 1

Syddansk Universitet, Mads Clausen Institute, NanoSYD, Alsion 2, DK-6400 Sønderborg, Denmark E-mail: [email protected] 2 University of Bonn, Kekulé-Institute of Organic Chemistry and Biochemistry, Gerhard-Domagk-Str. 1, D-53121 Bonn, Germany Abstract. The growth of 2,5-Di-4-biphenyl-thiophene (PPTPP) on the dielectric substrates NaCl, KCl, KAP, muscovite mica, and phlogopite mica is investigated by atomic force microscopy (AFM) and fluorescence microscopy. In all cases fibers are formed with several ten nanometers height and several hundred nanometers width, respectively. Only for PPTPP on muscovite mica the fibers are mutually parallel aligned along a single substrate direction, i.e. along muscovite 〈110〉. This uniaxial growth is explained by an electrostatic interaction between the molecules and surface electric fields in combination with epitaxy. The various growth directions on other substrates are dictated by epitaxy alone.

1.

Introduction

Nanofibers from organic conjugated molecules such as bare and functionalized para-phenylenes [1,2], α-thiophenes [3], phenylene-thiophene co-oligomers [4,5], or coumarin derivatives [6] have a high application potential in future optoelectronic devices [7]. Waveguiding [8–10], tunable light emission [11], gain narrowing and lasing [12–14], as well as frequency doubling [15,16] have already been observed. In the past the growth of the blue-light emitting para-hexaphenylene (p-6P) molecules on muscovite mica and on KCl has been investigated in detail [17–20]. For p-6P on muscovite mutually parallel aligned needles along a single 〈110〉 substrate direction have been observed, their growth being steered by the anisotropic electric surface fields [21]. This 〈110〉 direction corresponds to the direction of grooves on the muscovite surface, which alternate by 120° in between consecutive cleavage planes. This direction is denoted as 〈110〉g, the non-grooved one as 〈110〉ng. The long molecule axis is oriented at ±76° with respect to muscovite 〈110〉g, resulting in a single energetically favorable needle orientation. Low energy electron diffraction (LEED) and thermal desorption spectroscopy (TDS) have shown, that the first growth step is the formation of a wetting layer from lying molecules. Then clusters grow, and finally these clusters aggregate into needles. On KCl no such wetting layer and almost no clusters have been detected [22]. Needles grow along the two 〈110〉 substrate directions simultaneously [23]. For the investigated rod-like molecules so far the possible needle directions on muscovite depend on the epitaxial orientation of the molecules on the substrate, and on the packing of these molecules within a needle [24]. Therefore parafunctionalized para-phenylenes show a very similar growth behavior compared to p-6P. The α-thiophenes quaterthiophene and sexithiophene, however, align with their long molecular axes along the two muscovite high symmetry directions without H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_2, © Springer-Verlag Berlin Heidelberg 2009

11

12

F. Balzer et al.

a groove, i.e. along 〈110〉ng, and [100], and therefore three needle orientations are accomplished simultaneously. That way fortunate optical properties of an uniaxially aligned needle film such as polarized light absorption and emission are lost. To preserve some of the phenylene growth properties but, e.g., change the emission color substantially we therefore have investigated the growth of a thiophene phenylene co-oligomer: 2,5-Di-4-biphenyl-thiophene (PPTPP), Fig. 1. Here we show first results for growth experiments on different substrate surfaces under similar growth conditions, i.e. similar substrate temperatures Ts, deposition rates, and nominal film thicknesses.

2.

Experimental Methods

2,5-Di-4-biphenyl-thiophene has been synthesized in a two-fold Suzuki crosscoupling reaction from commercially available 2,5-dibromo-thiophene and 4-biphenyl boronic acid, using 5 mol% tetrakis(triphenylphosphino)palladium as catalyst together with cesium fluoride as base in dry tetrahydrofurane. The desired product has been obtained in yields of 80% after refluxing for 50 h. The final product precipitated from the reaction mixture and was washed with water and organic solvents repeatedly for purification. By outgassing in vacuo residual organic solvents are removed to give the desired compounds in high purity. Note that PPTPP has been synthesized previously using very similar approaches [25,26]. HO Br Br

+ 2

S

B HO

Pd(PPh3)4 + CsF THF, Δ, 50 h

Suzuki Cross Coupling Conditions

S 80 % yield

PPTPP citreous amorphous solid

Fig. 1. Schematics of the synthesis of 2,5-Di-4-biphenyl-thiophene (PPTPP).

As substrate materials five different single crystals have been chosen: potassium chloride (KCl), sodium chloride (NaCl), potassium acid phthalate (KAP) [27], muscovite mica [28], and phlogopite mica. All are cleaved in air and are transferred immediately into a high vacuum system (base pressure 2 × 10–8 mbar). Either right away or after annealing a clear low energy electron diffraction (MCPLEED, Omicron) pattern is observed for each of them. Organic molecules are deposited from an effusion cell. The nominal deposited film thickness is estimated by a water-cooled quartz microbalance located next to the substrate.

Organic Nanofibers from PPTPP

13

After deposition the samples are characterized in situ by LEED, ex situ by atomic force microscopy (AFM, JPK NanoWizard) and fluorescence microscopy (excitation wavelength λexc = 365 nm from a high-pressure mercury lamp).

3.

Results and Discussion

Figure 2 shows 150 × 150 µm2 fluorescence microscope images of typical samples deposited at Ts = 350 K – 380 K with a deposition rate of 0.1 Å/s – 0.2 Å/s and nominal thicknesses up to 5 nm. On all of those substrates needle-like structures from PPTPP form. All emit blue-green light after normal incidence UV irradiation. The emitted fluorescence from the needles is strongly polarized, the polarization vector being oriented perpendicular to the local needle direction. Similar to the case of, e.g., p-6P [21,29,30] this points to fibers made from lying organic molecules, the long molecular axis being perpendicular to the long needle axis. In addition for all substrates except for muscovite mica a green light emitting background is visible. This can be clearly seen in Fig. 2(c), where the border between the bare substrate and the deposition area is imaged. Typical fluorescence spectra are presented in Fig. 3. A well resolved vibronic progression is visible between 400 nm (a)

(b)

(d)

(c)

(e)

Fig. 2. Fluorescence microscope images, 150 × 150 µm2, of PPTPP on (a) NaCl, (b) KCl, (c) KAP, (d) muscovite mica, and (e) phlogopite mica. The nominal thickness of all samples is between 2 nm and 5 nm, the surface temperature during deposition varies between Ts = 350 K – 380 K.

14

F. Balzer et al.

Fig. 3. Unpolarized measured fluorescence spectra of PPTPP needle films on different substrates after excitation with unpolarized λexc = 365 nm light under normal incidence. The dotted vertical lines mark the peak positions at 437 nm, 463 nm, 493 nm, 529 nm, and 570 nm.

and 600 nm with an energy spacing of approximately 1300 cm–1, being most prominent in the case of muscovite and phlogopite mica. Slightly different colors in the microscopy images stem from different relative intensities of the excitonic transitions. The reason for the distribution of intensity between the fluorescence peaks is still unknown, but might be related to different molecule–molecule interactions for the molecules forming the patches of upright molecules, see below, and for molecules forming needles [31]. Although needle-like structures are formed in all five cases, the realized needle directions depend very much on the substrate. On NaCl, KCl, and KAP two needle directions evolve, with different angles in between. On NaCl and KCl the angle is 90°, and the needles grow along the two substrate 〈110〉 directions simultaneously. On KAP the angle is 68°, corresponding to the angle between the two KAP 〈101〉 directions [32]. On muscovite mica all of the needles grow along a single 〈110〉 direction, with altogether two orientational domains being present on the whole sample. On phlogopite mica three different growth directions exist, most pronounced at the very early growth stages and for deposition at room temperature. For larger coverages and higher substrate temperatures the needles tend to curl and to form rings. To a lesser extent this curling is also observed for PPTPP on NaCl and muscovite, Figs. 2(a) and (d), the tendency increasing with the overall needle length. Corresponding AFM images (see Fig. 4) provide widths and heights of the needles, but also show in more detail the greenish background from Fig. 3. This background results from patches of one or two multiples of approximately 2.2 nm height, suggesting that they are formed from upright molecules [25]. This background appears for all substrates except for muscovite mica, where clusters appear instead. These clusters are also found on phlogopite mica. They do not exist directly besides the needles (denuded zones).

Organic Nanofibers from PPTPP (a)

(b)

(c)

(d)

(e)

(f)

15

(g)

Fig. 4. Atomic force microscope images, 40 × 40 µm2, of PPTPP on (a) NaCl, (b) KCl, (c) KAP, (d) muscovite mica, and (e) phlogopite mica, corresponding to the fluorescence microscope images from Fig. 2. The height scales are (a) – (c) 100 nm, and (d) – (e) 50 nm. Arrows emphasize substrate directions. In the 5 × 2.5 µm2 AFM images for muscovite (f) and phlogopite (g) the clusters from lying molecules as well as layers from upright ones are clearly visible.

4.

Conclusions and Outlook

Obviously the growth directions of the needles depend strongly on the crystal structure of the underlying substrates. The symmetry of the substrate is retained in the needles’ growth directions. Similar to the case of p-6P we attribute this observation to an epitaxial relationship of the organic molecules with the corresponding substrate. On NaCl and on phlogopite the alignment is not as perfect as for KCl and muscovite mica, respectively. The uniaxial growth on muscovite mica is explained by an electrostatic interaction between the molecules and surface electric fields in combination with epitaxy. Epitaxy leads to an alignment of the molecules along muscovite high symmetry directions, whereas the electric fields choose the energetically most favorable of the three possible growth directions: needles grow along 〈110〉g. Muscovite and phlogopite mica exhibit almost identical lattice constants and surface compositions, but differ in that phlogopite is a trioctahedral mica, whereas muscovite is a dioctahedral one. This leads to the already describe grooves along a single 〈110〉 direction on muscovite, which are missing on phlogopite [33]. Therefore on phlogopite three simultaneous needle directions exist, on muscovite only one.

16

F. Balzer et al.

Not only the realized needle directions, but also the growth mechanism is similar to the case of p-6P. The observed clusters on muscovite and phlogopite are remnants from the initial growth stage. Only when the cluster number density reaches a critical value, needles start growing, mainly by agglomeration of the clusters. A LEED pattern from a wetting layer of lying molecules is observed for the case of muscovite mica. For all other cases no such diffraction pattern has been detected. As a conclusion the growth mechanism of PPTPP and the realized needle directions on the different substrates are similar to that of p-6P. Understanding such basic growth principles allows one to predict qualitatively nanowire surface growth from other conjugated molecules and thus allows for a sophisticated design of new devices. Adding, for example, another thiophene ring next to the existing one to form 5,5´-Di-4-biphenyl-2,2´-bithiophene (PPTTPP) does not alter the basic growth mode, but changes the epitaxial alignment on muscovite. From that two simultaneous needle directions on muscovite are predicted and are actually observed [24]. Acknowledgements. We are grateful to the Danish research agencies FNU and FTP as well as the Danish Advanced Technologies Trust for supporting this work by various grants. MS and AL thank the German research foundation DFG for financial support.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

F. Balzer and H.-G. Rubahn, Adv. Funct. Mater., 15, 17, 2005. M. Schiek, F. Balzer, K. Al-Shamery, A. Lützeu, and H.-G. Rubahn, Soft Matter, 4, 277, 2008. F. Balzer, L. Kankate, H. Niehus, and H.-G. Rubahn, Proc. SPIE, 5724, 285, 2005. H. Yanagi, T. Morikawa, S. Hotta, and K. Yase, Adv. Mater. 13, 313, 2001. F. Balzer, M. Schiek, A. Lützeu, K. Al-Shamery, and H.-G. Rubahn, Proc. SPIE 6470, 647006, 2007. M. Mille, J.-F. Lamere, F. Rodrigues, and S. Fery-Forgues, Langmuir 24, 2671, 2008. K. Al-Shamery, H.-G. Rubahn, and H. Sitter, editors. Organic Nanostructures for Next Generation Devices, Vol. 101 of Springer Series in Materials Science, Berlin 2008. K. Takazawa, Chem. Phys. Lett. 452, 168, 2008. H. Yanagi and T. Morikawa, Appl. Phys. Lett. 75, 187, 1999. F. Balzer, V.G. Bordo, A.C. Simonsen, and H.-G. Rubahn, Phys. Rev. B 67, 115408, 2003. Y.S. Zhao, H. Fu, F. Hu, A. Peng, W. Yang, and J. Yao, Adv. Mater. 20, 79, 2008. F. Quochi, F. Cordella, A. Mura, G. Bongiovanni, F. Balzer, and H.-G. Rubahn, J. Phys. Chem. B 109, 21690, 2005. F. Quochi, F. Cordella, R. Orru, J.E. Communal, P. Verzeroli, A. Mura, G. Bongiovanni, A. Andreev, H. Sitter, and N.S. Sariciftci, Appl. Phys. Lett. 84, 4454, 2004. H. Yanagi, T. Ohara, and T. Morikawa, Adv. Mater. 13, 1452, 2001. F. Balzer, J. Brewer, J. Kjelstrup-Hansen, M. Madsen, M. Schiek, K. Al-Shamery, A. Lützeu, and H.-G. Rubahn, Proc. SPIE 6779, 67790I, 2007.

Organic Nanofibers from PPTPP

17

16 J. Brewer, M. Schiek, A. Lützeu, K. Al-Shamery, and H.-G. Rubahn, Nano Lett. 6, 2656, 2006. 17 L. Kankate, F. Balzer, H. Niehus, and H.-G. Rubahn, J. Chem. Phys. 128, 084709, 2008. 18 A. Andreev, T. Haber, D.-M. Smilgies, R. Resel, H. Sitter, N.S. Sariciftci, and L. Valek, J. Cryst. Growth 275, e2037, 2005. 19 A.Y. Andreev, C. Teichert, G. Hlawacek, H. Hoppe, R. Resel, D.-M. Smilgies, H. Sitter, and N.S. Sariciftci, Org. Electron. 5, 23, 2004. 20 P. Frank, G. Hlawacek, O. Lengyel, A. Satka, C. Teichert, R. Resel, and A. Winkler, Surf. Sci. 601, 2152, 2007. 21 F. Balzer and H.-G. Rubahn, Appl. Phys. Lett. 79, 3860, 2001. 22 P. Frank, G. Hernandez-Sosa, H. Sitter, and A. Winkler, Thin Solid Films 516, 2939, 2008. 23 T. Mikami and H. Yanagi, Appl. Phys. Lett. 73, 563, 1998. 24 F. Balzer, M. Schiek, K. Al-Shamery, A. Lützeu, and H.-G. Rubahn, J. Vac. Sci. Technol. B 26, 2008. In print. 25 T.J. Dingemans, N.S. Murthy, and E.T. Samulski, J. Phys. Chem. B 105, 8845, 2001. 26 S. Hotta, H. Kimura, S.A. Lee, and T. Tamaki, J. Heterocycl. Chem. 37, 281, 2000. 27 M. Campione, A. Sassella, M. Moret, A. Papagni, S. Trabattoni, R. Resel, O. Lengyel, V. Marcon, and G. Raos, J. Am. Chem. Soc. 128, 13378, 2006. 28 E.W. Radoslovich, Acta Cryst. 13, 919, 1960. 29 A. Andreev, G. Matt, C.J. Brabec, H. Sitter, D. Badt, H. Seyringer, and N.S. Sariciftci, Adv. Mater. 12, 629, 2000. 30 A. Niko, E. Zojer, F. Meghdadi, C. Ambrosch-Draxl, and G. Leising, Synth. Met. 101, 662, 1999. 31 E. Da Como, M.A. Loi, M. Murgia, R. Zamboni, and M. Muccini, J. Am. Chem. Soc. 128, 4277, 2006. 32 A.V. Alex and J. Philip, J. Appl. Phys. 88, 2349, 2000. 33 Y. Kuwahara, Phys. Chem. Miner. 28, 1, 2001.

α-Sexithiophene Films Grown on Cu(110)-(2x1)O: From Monolayer to Multilayers Martin Oehzelt1,2, Stephen Berkebile2, Georg Koller2, Thomas Haber2, Markus Koini3, Oliver Werzer3, Roland Resel3, and Michael G. Ramsey2 1

Institute of Experimental Physics, Johannes Kepler University Linz, Altenbergerstraße 69, A-4040 Linz, Austria E-mail: [email protected] 2 Institute of Physics, Karl-Franzens University Graz, Universitätsplatz 5, A-8010 Graz, Austria 3 Institute of Solid State Physics, Graz University of Technology, Petersgasse 16, A-8010 Graz, Austria Abstract. The growth of α-sexithiophene (6T) on copper (110) and oxygen reconstructed Cu(110) is studied by multiple techniques such as STM (scanning tunnelling microscopy), XRD (X-ray diffraction), XPS (X-ray photoelectron spectroscopy) and NEXAFS (near edge X-ray absorption fine structure). Selected data will be presented here and we will show that the long axes of the molecules on Cu(110) and Cu(110)-(2x1)O (CuO) are aligned along the valleys of the surface corrugations, i.e. along [1–10] and [001], respectively. With GIXD (grazing incidence X-ray diffraction) measurements the monolayer structure of 6T on Cu-O could be determined. Thicker films were studied by the X-ray diffraction pole figure technique. On all surfaces the (010) net planes of the bulk crystal structure are parallel to the surface i.e. the films grow exclusively (on Cu-O) or pre- dominantly (on Cu) in the (010) orientation. In the thick film the long molecular axes of the 6T molecules are found to be parallel to those of the monolayer. To study the transition from the monolayer to the multilayer structures NEXAFS measurements were carried out.

1.

Introduction

Organic semiconductors have attracted considerable interest in the last decades and as a result remarkable progress in the field of organic electronics has been made [1]. Especially their potential use in low-cost devices such as organic solar cells, light emitting diodes or thin film transistors have encouraged many investigations including basic research studies [2]. Here we present data on the growth of α-sexithiophene (6T), as this molecule is one of the most prominent organic transistor materials [3] and the focus will be on structural changes of 6T on the Cu(110)-(2x1)O surface from the initial nucleation to thick films. Additional information can be found in the following papers [4–7].

2.

Results and Discussion

The STM image of Fig. 1a reveals the molecular orientation of the 6T molecules in the monolayer regime. The long molecular axis is parallel to the substrate H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_3, © Springer-Verlag Berlin Heidelberg 2009

19

20

M. Oehzelt et al.

surface and parallel to the Cu-O rows which is the Cu[001] direction. The molecules are stacked in rows perpendicular to the long molecular axis. While the order between the stacks is less pronounced, the distance between neighboring molecules is 5.1 Å and commensurate to the Cu-O rows. This ordering is also visible in the Fourier transform of the image – sharp points for the 5.1 Å periodicity and streaks in the perpendicular direction.

Fig. 1. STM images of a monolayercoverage of 6T on Cu(110)-(2x1)O. Tunneling parameter: Vt = +2.0V, It = 1nA. The high symmetry directions are indicated and the long molecular axes are parallel to [001]. The insert is the fourier transformation of the STM image which reveals the characteristic distances between the molecules. Figure 1b is a GIXD pattern for net-planes perpendicular to the long molecular axis and reflecting the periodicity between molecular stacks in the monolayer regime of 27 Å.

For thicker films deposited on Cu-O, the molecules crystallize in their bulk crystal structure [4] with 6T(010) net-planes parallel to the substrate. The in-plane orientation is such that the long molecular axes are still aligned along the Cu-O rows [4] and therefore they clearly adopt the molecular orientation of the monolayer. Note that in the 6T(010) net-plane the molecules are still organized in stacks, but in contrast to the monolayer this surface unit cell is oblique and the stacking direction is not 90° with respect to the long molecular axis but 66.5°. The thickness where this transition from a parallel stacking to a oblique one takes place is still subject of further investigations. Nevertheless, recent grazing incidence X-ray diffraction (GIXD) experiments have shown that 6T films with bulk crystallites still show the ordered monolayer structure underneath. In Fig. 1b a GIXD graph is shown where a series of peaks originating from the monolayer are shown. The distances determine the spacing between the stacks and picking up the 27 Å periodicity of the long molecular axis. Figure 2 shows the results from a series of NEXAFS measurements. Details about the measurement geometry and the data evaluation can be found in [4] and references therein. In summery the NEXAFS measurements reveal that the molecular tilt angle in the monolayer is 4° higher than in the multilayer. This result combined with the knowledge that the monolayer structure with a spacing of 5.1 Å is reduced compared to the multilayer one leads to the following model: The spacing between adjacent molecules determined by the Cu-O rows ends up

α-Sexithiophene Films Grown on Cu(110)-(2x1)O: From Monolayer to Multilayers 21

in a higher tilt angle to keep their Van der Waals distance (d) constant (see Fig. 2). With increasing thickness the bulk structure develops. How the different stacking within the molecular rows comes into play and how rapidly the transition from the monolayer to the multilayer occurs are still open questions.

Fig. 2. NEXAFS spectra for a thick 6T layer (left) and the monolayer (right) grown on Cu-O are shown (for details see also Ref. 5). The model in the middle shows the change in the molecular packing according to the NEXAFS, STM and XRD measurements.

Acknowledgements. This work was supported by the Austrian Science Foundation (FWF) and by the European Synchrotron Research Facility (ESRF).

References 1 2 3 4 5 6 7

N. Koch in ChemPhysChem, Vol. 8, 1438, 2007. A. Facchetti, M.H. Yoon, J.T. Marks in Advanced Materials, Vol. 17, 1705, 2005. H.E. Katz in Journal of Materials Chemistry, Vol. 7, 376, 1997. M. Oehzelt, G. Koller, J. Ivanco, S. Berkebile, T. Haber, R. Resel, F.P. Netzer, M.G. Ramsey in Advanced Materials, Vol. 18, 2466, 2006. M. Oehzelt, L. Grill, S. Berkebile, G. Koller, F.P. Netzer, M.G. Ramsey in ChemPhysChem, Vol. 8, 1707, 2007. M. Oehzelt, S. Berkebile, G. Koller, J. Ivanco, S. Surnev, M.G. Ramsey, in preparation. M. Koini, T. Haber, O. Werzer, S. Berkebile, G. Koller, M. Oehzelt, M.G. Ramsey, R. Resel, in preparation.

Para-Sexiphenyl Layers Grown on Light Sensitive Polymer Substrates G. Hernandez-Sosa1, C. Simbrunner1, T. Höfler2, A. Moser3, O. Werzer3, B. Kunert3, G. Trimmel2, W. Kern2,4, R. Resel3 and H. Sitter1 1

Institute for Semiconductors and Solid State Physics, Johannes Kepler University Linz, Altenbergerstrasse 69, 4040 Linz, Austria 2 Institute for Chemistry and Technology of Materials, Graz Technical University, Stremayrgasse 16, 8010 Graz, Austria 3 Institute of Solid State Physics, Graz Technical University, 8010 Graz, Austria, Petergasse 16, 8010 Graz, Austria 4 Department of Chemistry of Polymeric Materials, Montanuniversität Leoben, Franz-Josef-Strasse 18, 8700 Leoben, Austria Corresponding author E-mail: [email protected] Abstract. In this contribution the deposition of Para-sexiphenyl (PSP) layers on poly (diphenyl bicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate) (PPNB) by Hot Wall Epitaxy (HWE) is reported. It is demonstrated that pre-treating the substrate by UV-illumination induces a clear change in the morphology of the grown PSP films due to the polarity modification of the substrate surface. PPNB surface polarity increases when illuminated by UV via photo-Fries rearrangement. By detailed atomic force microscopy analysis the influence on the growth kinetics by the substrate temperature, deposition time and particularly by the UVtreatment of the substrate was investigated. A high crystalline order of the films is underlined by the observation of growth spirals and terraced islands, providing mono-layer step heights of standing PSP-molecules.

1.

Introduction

Morphology and crystalline order is determining for improving the electrical and optical properties of organic films. Therefore it is of great importance to study the growth kinetics of organic materials on substrate surfaces with well controlled properties. Para-sexiphenyl (PSP) (C36H26), a six units oligomer of para-phenylene, is a promising candidate as an electro active layer in organic LED displays due to its blue luminescence with high quantum yield. [1,2]. Moreover, it is classified as a wide gap organic semiconductor with an electronic band gap of 3.1 eV with photoluminescence in the blue visible range and polarized absorption and emission when provided in well ordered films [3]. The Hot-wall epitaxy (HWE) technique offers the advantage that it works close to thermodynamic equilibrium. Therefore it allows organic molecules to find the most suitable arrangement into the crystal lattice and as a consequence, highly ordered organic thin films can be obtained [3,4]. Organic thin films grown by HWE have shown outstanding optical and electrical properties [5,6]. Extensive morphological and structural characterization has been already performed on HWE grown

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_4, © Springer-Verlag Berlin Heidelberg 2009

23

24

G. Hernandez-Sosa et al.

PSP films deposited on various substrates such as KCl and muscovite mica, showing that the nature of the substrate and the growth conditions are ruling parameters for the molecular packing of the films [7]. In the presented work, an amorphous polymer poly(diphenyl bicyclo[2.2.1] hept-5-ene-2,3-dicarboxylate) (PPNB) containing photoreactive aryl-ester groups was chosen as substrate. Upon illumination with UV-light of λ < 280 nm these ester groups isomerise to the corresponding hydroxyketones in the so-called photo-Fries reaction [8,9]. Recently, we have shown that the photoreaction in PPNB yields up to 21% hydroxyketones as photoproducts [10]. The UV-induced reaction leads to a large increase of the refractive index as well as of the surface polarity. This enhanced polarity and the generated new functional groups can be used for selective post-modification reactions [11]. In the present work the influence of this change of surface polarity on the surface morphology of the PSP deposited layers is studied.

2.

Experimental Methods

The procedure to synthesise PPNB is reported by T. Höfler et al. [10]. For the substrate preparation a 10 mg/ml solution of PPNB in CHCl3 was prepared and stirred for 12 h. Then the solution was spin cast onto Si-substrates resulting in 80 nm film thickness. In order to provide equivalent growth conditions for UVexposed and non-irradiated surfaces, each substrate was divided in two regions. One half of the substrate was exposed for 20 min to UV light (254 nm) while the other half was protected from UV-illumination. The unfiltered light of an ozonefree mercury low pressure UV lamp (Heraeus Noblelight; 254 nm) was used. The illumination-process was done in inert gas atmosphere (nitrogen with a purity >99.95%) in order to avoid unwanted oxidation reactions. After the illumination process the substrates were transferred via a load lock to a HWE growth chamber working at a dynamic vacuum of 9 × 10–6 mbar. A 15 min in-situ preheating procedure was applied in order to reduce surface contaminations. The substrate temperature during preheating is chosen the same as the growth temperature in order to allow constant thermal conditions during the whole deposition process. The preheating process also removes possible adsorbed species from the surface of the substrate. A complete series of samples was prepared varying growth time (5–60 min.) and substrate temperature (100°C – 160°C) whereas the source- (240°C) and wall-temperature (260°C) were kept constant. The working principles of a HWE system can be found elsewhere [12]. Ex-situ atomic force microscopy (AFM) studies were performed using a Digital Instruments Dimension 3100 in the tapping mode. A SiC tip was used and the scanning area was 10 × 10 μm.

3.

Results and Discussion

PSP was deposited using the HWE technique on PPNB at different preparation conditions as explained in the experimental methods section. A detailed morphology analysis of the grown sample series, using AFM in tapping mode was performed.

Para-Sexiphenyl Layers Grown on Light Sensitive Polymer Substrates

25

Fig. 1. Atomic Force Microscopy images (10 μm × 10 μm) of Para-sexiphenyl on PPNB previously treated with UV light (upper half of each image) and as prepared (bottom half) for different substrate temperatures and different deposition times.

The influence of the polymer substrate on the surface properties of the grown PSP is analysed depending on substrate temperature, deposition time and UVillumination of the substrate before growth. Figure 1 depicts a chart with the AFM images for grown PSP films at, 100°C, 130°C and 160°C varying the deposition time from 5 to 60 min. The morphology of the film grown on the UV-illuminated side and on the as prepared surface are on the upper and bottom side of each image, respectively. The height scale (z0) is presented at the bottom of each image. A clear morphological difference between non-illuminated and UV-illuminated regions can be observed. Whereas on the non-illuminated side a homogeneous PSP film is formed, pre treating the substrate by UV illumination is leading to island formation and consequently to a change of the growth kinetic. This behaviour is a direct consequence of the increase in surface polarity from the substrate resulting of the photochemical reaction of PPNB described by T. Höfler et al. [10]. Changing the deposition time of PSP from 5 to 60 min is leading to an areaexpansion of the single islands. A more detailed comparison of the as prepared and UV-illuminated regions of the sample grown for 20 min at 160°C (Fig. 1-IIIb) is presented in Figure 2. On the one hand on the as prepared substrate (Fig. 2a, Fig. 1-IIIb bottom half) we can observe a closed film with spiral features which are characteristic of screw dislocations. On the other hand, on the UV illuminated side of the substrate (Fig. 2c, Fig. 1-IIIb upper half) crystallites with terraces of up to hundredths of nanometres in length can be observed, underling a change in the growth procedure.

26

G. Hernandez-Sosa et al.

Fig. 2. Atomic force microscopy scans and profiles of the surface structures formed by PSP deposited on the as prepared (a,b) and pre UV illuminated (c,d) PPNB surface. Scan size is 2 × 2 μm.

Figures 2b,d represent a detailed analysis of cross sections indicated in figures 2a,c as black solid lines. It is demonstrated that the step heights of the observed features are in good agreement with the value for one monolayer of standing PSP molecules corresponding to 2.6 nm. The formation and increment in size of these crystallites is induced by increasing the substrate temperature during PSP deposition, which is observed for both – illuminated and non-illuminated substrates – as shown in Figure 1. Similar results are found in literature for small organic molecules [13].

4.

Conclusions

High quality Para-Sexiphenyl films where successfully deposited on PPNB at different temperatures and deposition times. Careful AFM investigations show that PSP shows a different morphology and a different growth process when deposited on UV illuminated or as prepared PPNB substrates. This effect is attributed to the change in surface polarity after the UV illumination treatment and to the fact that the surface becomes more hydrophilic. Both changes are consequence of the photo-Fries rearrangement that the aryl-ester groups of PPNB undergo upon UV irradiation. On this work it is demonstrated that, the surface morphology of PSP grown on PPNB can be influenced by a simple and well controlled pre-treatment of the substrate surface and deposition conditions. This procedure could potentially influence the optical and electrical properties of the films deposited on the different parts of the substrate. Therefore the demonstrated sample structure opens new perspectives for the fabrication of devices which electrical and optical properties can be controlled by UV treatment of the substrate.

Para-Sexiphenyl Layers Grown on Light Sensitive Polymer Substrates

27

Acknowledgements. This work was supported by the Austrian Science Foundation projects NFN-S9702, NFN-S9706 and NFN-S9708. GHS wants to thank Consejo Nacional de Ciencia Tecnología (CONACYT), in México for scholarship.

References 1 2

3 4 5

6

7 8 9 10 11 12 13

S. Tasch, C. Brandstatter, F. Meghdadi, G. Leising, G. Froyer, L. Athouel, Advanced Materials, Vol. 9, 33, 1997. G. Leising, S. Tasch, W. Graupner, Fundamentals of Electroluminescence in Paraphenylene-type Conjugated Polymers and Oligomers, in Handbook of Conducting Polymers, Edited by T. Skothem, R. Elsenbaumer, J. Reynolds, 2nd ed., New York 1997. A.Y. Andreev, G. Matt, C.J. Brabec, H. Sitter, D. Badt, H. Seyringer, N.S. Sariciftci, Advanced Materials ,Vol. 12, 629, 2000. D. Stifter and H. Sitter, Applied Physics Letters, 66, 679, 1995. Th. B. Singh, N. Marjanovic, G.J. Matt, S. Gunes, N.S. Sariciftci, A.M. Ramil, A. Andreev, H. Sitter, R. Schwodiauer, and S. Bauer, Organic. Electronics. Vol. 6, 105, 2005. F. Quochi, F. Cordella, R. Orru, J.E. Communal, P. Verzeroli, A. Mura, and G. Bongiovanni A. Andreev, H. Sitter, and N.S. Sariciftci Applied Physics Letters, Vol. 84, 4454, 2004. T. Haber, A. Andreev, A. Thierry, H. Sitter, M. Oehzelt, R. Resel, Journal of Crystal Growth, Vol. 284, 209, 2005. J.C. Anderson, C.B. Reese, Proceedings of the Chemical. Society, 217, 1960. D. Bellus, Advances in Photochemistry, Vol. 8, 109, 1981. T. Höfler, T. Griesser, X. Gstrein, G. Trimmel, G. Jakopic, W. Kern, Polymer, Vol. 48, 1930, 2007. T. Griesser, T. Höfler, S. Temmel, W. Kern, G. Trimmel, Chemistry of Materials, Vol. 19, 3011, 2007 A. Lopez-Otero, Thin Solid Films, Vol. 3, 4, 1978. Th.B. Singh, N.S. Sariciftci, H. Yang, L. Yang, B. Plochberger and H. Sitter, Applied Physics Letters, Vol. 90, 213512, 2007.

Thermal Desorption of Organic Molecules Adolf Winkler Institute of Solid State Physics, Graz University of Technology, Petersgasse 16, A-8010 Graz, Austria E-mail: [email protected] Abstract. The applicability of thermal desorption spectroscopy (TDS) to differentiate between a strongly bound wetting layer and the subsequently formed multilayers of organic thin film is discussed. After describing the fundamentals of TDS, particularly for large organic molecules, several model systems are presented (p-4P on Au(111), p-6P on Au(111), mica(0001) and KCl(001)), demonstrating the importance of the substrate material, surface structure and surface contaminations on the formation of a wetting layer. The wetting layer acting as a template for the multilayer growth strongly influences the structure and morphology of the organic thin film.

1.

Introduction

Thermal desorption spectroscopy (TDS) is a powerful and well known experimental technique to investigate adsorption/desorption kinetics and ener-getics of small (inorganic) molecules on surfaces [1,2]. The application of this technique for large organic molecules and ultra-thin organic films is not that widely acknowledged. In this review I will demonstrate that TDS can be successfully used to obtain information on a number of parameters characterizing organic films in the sub-monolayer, monolayer and near multilayer regime. In particular, the frequently put question as to the existence of a wetting layer prior to the growth of island like films can be answered unambiguously. In addition, the binding energy of the first monolayer and the heat of evaporation of the multilayer can be determined. Other kinetic parameters of adsorption and desorption, like the sticking coefficient and the pre-exponential factor for desorption can be obtained as well. In this contribution I will exemplify the power of TDS on some model systems: p-quaterphenyl (p-4P) on Au(111) and p-hexaphenyl (p-6P) on Au(111), mica(0001) and KCl(001). I will particularly show that the chemical composition (carbon contamination) and the morphology (surface roughening by sputtering) of the substrate surface can have a tremendous effect on the thin film growth.

2.

Fundamentals of TDS

For thermal desorption spectroscopy (TDS) the material is first put onto the surface and thereafter the sample is heated with a constant heating rate until the material is desorbed again from the surface. The desorbing particles are detected with a mass spectrometer and the signal vs. time (or temperature) is called the TD H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_5, © Springer-Verlag Berlin Heidelberg 2009

29

30

A. Winkler

spectrum. In most cases the desorption spectrum can be successfully described by the Polanyi-Wigner equation:

dN dN x R =− =− β = ν ⋅ N ⋅ exp( − E (Θ)/ kT ) des des dt dT

(1)

here is R des : desorption rate, β: heating rate, ν: pre-exponential factor, N: number of molecules per square unit, x: desorption order, Edes: desorption energy, Θ: coverage, T: surface temperature, k: Boltzmann’s constant. The coverage Θ is usually defined (for small inorganic molecules and atoms) as the ratio of the number of adsorbed particles to the number of surface atoms per square unit. It is given in monolayers (ML), where the saturation coverage is typically 1 ML. For large organic molecules the value of the saturation coverage would be much smaller than unity. In this case the coverage is frequently defined as the ratio of the number of adsorbed molecules to the maximum number of adsorbed molecules per square unit (Nmax), which can be regarded as one physical monolayer. Then N = Nmax·Θ. The desorption order x describes the coverage dependence of the desorption rate. For organic molecules only the desorption orders x = 1 (first order) and x = 0 (zero order) are relevant. Zero order desorption takes place in the presence of multilayers. In this case the maximum number of molecules per surface unit is always available and therefore no coverage dependence exists (N0). When it comes to desorption of the last monolayer the desorption rate is proportional to the number of available molecules, i.e. proportional to N1. The desorption energy Edes is constant for multilayer desorption and is equal to the heat of evaporation. For the desorption of the final monolayer the desorption energy may be coverage dependent: E(Θ) = E0 ± ωΘ. The ± sign describes an attractive (+) or repulsive (−) lateral interaction between the molecules, with the interaction energy ω. The pre-exponential factor ν can be correlated with the attempt frequency of the adsorbed molecules to overcome the adsorption potential. For atoms and small molecules this value is typically in the order of 1013 s–1. For molecules with a large number of atoms this perception is not appropriate. The pre-exponential factor actually takes into account the change of all translational and internal degrees of freedom during desorption. As a result of transition state theory (TST) considerations [3] the pre-exponential factor can be described as:

⎛ kT ⎞ q⊕ ⎟ ⎝ h ⎠ q

ν =⎜

(2)

with h: Planck’s constant, q: partition function of the adsorbed state, q ⊕: partition function of the desorbed (free) state. For atoms and small molecules both partition functions are similar and therefore ν ≈ kT/h ≈ 1013 s–1, for typical desorption temperatures. For large molecules, however, the partition function of the free molecules is due to the many rotational and vibrational degrees of freedom much larger than for the adsorbed molecule, where only frustrated rotations and vibrations exist. Therefore ν is in this case typically by many orders of magnitude larger than 1013 s–1 [4,5].

Thermal Desorption of Organic Molecules

31

The evaluation of the desorption energy for multilayer desorption, i.e. the heat of evaporation, is straightforward. The plot of ln Rdes vs. 1/T for the leading edge of the spectrum yields a straight line, where the slope is equal to –Edes/k. If the desorption rate is known quantitatively then the intercept of this straight line with the Y-axis yields the pre-exponential factor. For first order desorption a characteristic feature of the spectra is the coverage independence of the peak maxima. According to Redhead [6] the correlation between the desorption energy Edes and the peak maximum Tm is given by:

Edes ⎛ ν =⎜ 2 kTm ⎝ β

⎞ ⎟ exp (− Edes / kTm ) ⎠

(3)

An approximate solution of this equation is:

⎛ ⎛νT ⎞ ⎞ Edes ≈ kTm ⎜ ln ⎜ m ⎟ − 3.64 ⎟ ⎝ ⎝ β ⎠ ⎠

(4)

For many inorganic adsorbates, where ν ≈ 1013 s–1, the following numerical equation holds (Redhead equation): Edes (cal / mol ) ≈ 60 ⋅ Tm ( K )

(5)

however, this approximation cannot be used for large organic molecules. A better approach is to determine first the pre-exponential factor ν for multilayer desorption and to take this value then for the desorption of the monolayer. For example, p-6P multilayer desorption from Au(111) yields a pre-exponential factor ν = 5.6 × 1025 s–1 [7]. This leads to a numerical equation (5) which contains a value of 124 instead of 60! This is one of the reasons why the extraction of desorption energies from TD spectra via the Redhead formula was sometimes performed incorrectly [8]. The shape of the desorption spectra and the shift of the peak maxima Tm with changing coverage also gives some valuable information. Zero order desorption is characterized by a sharp cutoff at the trailing edge when the coverage goes to zero (Fig. 1a). The peak maxima shift to higher temperatures with increasing coverage. However, this shape is never observed for real situations, because desorption of the last monolayer will always change to a first order reaction, which leads to a less sharp trailing edge (Fig. 1b). First order desorption is characterized by asymmetric peaks but with coverage independent peak maxima Tm (Fig. 1c). If the desorption energy is coverage dependent then a shift of the peak maxima as a function of coverage appears. In the case of repulsive interaction, as frequently observed in the sub-monolayer range for organic molecules, Tm shifts to lower temperature (Fig. 1d), whereas for attractive interaction a shift to higher temperature is observed with increasing coverage.

A. Winkler

Desorption rate / arbitary units

4.5 4

3.5

a

Desorption rate / arbitary units

32

3.5 3

2.5 2

1.5 1

0.5

Desorption rate / arbitary units

3.5 3

320

340

360

380

Temperature / K

2.5 2

1.5 1

0.5 0

250

2 1.5 1 0.5

2

c

300

350

Temperature / K

400

b

2.5

0

400

Desorption rate / arbitary units

0

3

320

340

360

380

Temperature / K

400

d

1.5

1

0.5

0 250

300

350

Temperature / K

400

Fig. 1. Calculated desorption spectra for pure zero-order desorption (a), for changing the desorption order from zero to first order at 1 monolayer (b), for pure first order desorption (c) and for first order desorption with repulsive lateral interaction (d).

3.

Experimental Technique

The technique of TDS is quite simple and straightforward. The molecules which have first been put onto the surface at sufficiently low surface temperature by gas dosing or evaporation are afterwards desorbed again by heating of the sample. Typically, linear heating rates between 1 to 10 K/s are applied. The desorbed material can be detected by a pressure gauge or by a mass spectrometer. Since the amount of material detected is quite small, TDS is generally performed under ultra-high vacuum conditions. According to Redhead [6] the desorption rate is correlated with the detected signal in the following way: Rdes = KSp / A + KVdp / dt

(6)

with S: pumping speed, p: pressure signal, V: volume of the vacuum chamber, A: surface area, K = 3.27 × 1019 molecules/l. In the case of very large pumping speed S the second term in Equ. 6 can be neglected and the desorption rate becomes directly proportional to the measured signal. Whereas this condition for small (volatile) gas molecules is not necessarily fulfilled in all cases, for condensable species like large organic molecules this is fulfilled to a great extent. In other words, condensable desorbing molecules do not

Thermal Desorption of Organic Molecules

33

contribute to an increased background pressure, but are immediately pumped away when hitting the surface of the vacuum chamber at room temperature (S→∞). But this also means that the particles have to be detected by an in-line-of-sight detector. Multiplexing of the mass spectrometer allows also to detect possible reaction products of the adsorbate. For example, the dehydrogenation of a p-6P monolayer on gold surfaces could be verified by this method [9]. On the other hand cracking of the molecules in the ionization region of the mass spectrometer has to be taken into account. In particular for large reactive organic molecules, e.g. for thiol based molecules, which are typically used for SAM preparation, the discrimination between the cracking of the molecule at the surface and in the mass spectrometer can be a challenging task [10,11]. Thermal desorption spectroscopy is a very sensitive method regarding the kinetics and energetics of the adsorbate, which in turn depends strongly on the surface conditions. Therefore, a comprehensive characterization of the surface prior to TDS is indispensable. Typically, Auger electron spectroscopy (AES) or X-ray photoelectron spectroscopy (XPS) are applied for the determination of the surface chemical composition and Low Energy Electron Diffraction (LEED) for the geometrical characterization.

4.

Wetting Layers Observed by TDS

TDS is particularly suited to differentiate between a strongly bound wetting layer of organic molecules and the more weakly bound molecules in the multilayer. In most cases rod or disk like molecules will be adsorbed in a flat-on configuration in the wetting layer. It turns out that the structural configuration of the wetting layer generally determines the structure and morphology of the subsequently formed multilayer [12,13]. The wetting layer depends strongly on the substrate material, the substrate structure and contaminations on the substrate surface. In Fig. 2 desorption spectra of p-6P on mica(0001) and KCl(001) are shown. On mica (Fig. 2a) the existence of a strongly bound wetting layer is clearly observed (ß-peak), which saturates at about 0.2 nm mean thickness [14]. With increasing coverage the less strongly bond α peak appears which does not saturate, which is representative for multilayer desorption. Thus, the needle like islands observed for this system grow on top of the wetting layer. In contrast, on the potassium chloride surface (Fig. 2b) no evidence for a strongly bound first layer can be seen [15]. Even for the smallest coverage of 0.1 nm the desorption peak is located at the leading edge of the multilayer peaks. Since it is known that for small coverage the film grows in the form of needle like islands [16,17] one can conclude that no wetting layer exists between the islands. The influence of carbon contamination on the p-6P layer growth on Au(111) is shown in Fig. 3. On the clean surface several strongly bound adsorption states (ß1, ß2, ß3) can be identified for small coverage (Fig. 3a). The ß1 and ß2 states have been explained to be due to flat lying and side tilted molecules with coverage of 0.5 monolayers each. The ß3 peak stems from a second layer which can still be energetically distinguished from the multilayer peaks denoted by α (not shown here) [7]. The contamination of the surface by carbon (e.g. by X-ray irradiation of a thick p-6P film) leads to a significant change of the wetting layer.

34

A. Winkler

Fig. 2. (a) TDS for p-6P from mica(0001) showing the wetting layer (ß-peak) and multilayer (α-peak) (After Frank et al. [14]), (b) TDS for p-6P from KCl(001) reveals that no wetting layer exists in this case (After Frank et al. [15]).

Only one rather weakly bound desorption state (β) shows up in this case before multilayer desorption starts (Fig. 3b). Also the second layer cannot be distinguished any longer from the multilayer peaks. The morphology of thicker films grown on these two substrates are totally different. Whereas in the former case needle like island are formed consisting of molecules oriented parallel to the surface, in the latter case mound like island consisting of standing molecules can be observed [18]. Similarly, on mica(0001) C-contamination or roughening of the surface by heavy sputtering leads to a removal of the wetting layer. This results in a film growth with mound like islands consisting of standing molecules [14].

Fig. 3. (a) TDS of p-6P from clean Au(111) shows three states (ß1-ß3) belonging to the wetting layer, (b) TDS of p-6P from carbon contaminated Au(111) shows only one state of a more weakly bound wetting layer (After Müllegger and Winkler [7]).

Thermal Desorption of Organic Molecules

5.

35

Determination of Desorption Energies for Wetting and Multilayers

The desorption energy of multilayers (heat of evaporation) can be easily obtained as outlined in section 2. As an example we show the multilayer desorption of p-4P from a carbon covered polycrystalline Au-foil (Fig. 4a) and the corresponding lnR vs. 1/T plot (Fig. 4b). In Fig. 4a one can clearly see the zero order desorption behavior as described above (common leading edge, sharp cutoff of trailing edge). The determination of the desorption energy from Fig. 4b yields Edes = 1.6 eV. For the pre-exponential factor one obtains ν = 5 × 1021 s–1 [19]. 4P/0.5ML C/Au(poly) 300 k

TDS - signal (arb. units)

α

film thickness 12 nm 6.0 nm 2.5 nm 1.0 nm

34

In(Rate)

a

p-4P on Au(poly) multilayer

b

32

30

m = 306

350

400 450 Temprature (K)

0.0025

500

0.0026

0.0027

0.0028

1/T

Fig. 4 (a) Series of TDS for p-4P from a polycrystalline gold surface as a function of mean thickness, (b) lnR vs. 1/T plot for determination of the desorption energy (After Müllegger et al. [19]).

The desorption energies for the monolayer states (wetting layer) can be calculated by using Equ. 4 and inserting the pre-exponential factor obtained from the multilayer peak. In Table 1 the desorption energies and pre-exponential factors for various adsorption systems are compiled. Table 1. Compilation of pre-exponential factors and desorption energies for wetting layers and multilayers of p-4P and p-6P on various substrates.

System

ν (multi) [s–1]

Edes(multi) [eV]

Edes(mono)-β2 [eV]

Edes(mono)-β1 [eV]

p-4P/Au(poly) p-4P/Au(111) p-6P/Au(111) p-6P/Mica(0001) p-6P/KCl(001)

5 × 1021 1.6 × 1021 5.6 × 1025 3.7 × 1025 3 × 1024

1.6 1.5 2.4 2.5 2.3

1.9 2.1 3.2 – –

2.9 2.6 3.6 2.9 –

36

A. Winkler

6.

Conclusion

Thermal desorption spectroscopy is a powerful method to identify and characterize wetting layers in ultra-thin organic films. Furthermore, TDS allows to determine energetic (desorption energy) and kinetic (pre-exponential factor, sticking coefficient) parameters for the individual adsorption systems. In combination with other surface analytical tools a quite comprehensive characterization of the initial stages of organic film growth can be obtained. This has been exemplified for oligophenylenes (p-4P, p-6P) on various substrates (Au(111), mica(0001) and KCl(001)). Acknowledgements. I would like to thank S. Müllegger and P. Frank for their valuable contributions to this work. The financial support by the Austrian Science Fund (FWF) is gratefully acknowledged.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

D. A. King, Surface Sci. 47, 384, 1975. A. M. de Jong, J. W. Niemantsverdriet, Surface Sci. 233, 355, 1990. V. P. Zhdanov, Surf. Sci. Rep. 12, 183, 1991. K. R. Paserba and A. J. Gellmann, Phys. Rev. Lett. 86, 4338, 2001. K. A. Fichthorn and R. A. Miron, Phys. Rev. Lett. 89, 196103, 2002. P. A. Redhead, Vacuum 12, 203, 1962. S. Müllegger and A. Winkler, Surface Sci. 600, 1290, 2006. C. B. France and B. A. Parkinson, Appl. Phys. Lett. 82, 1194, 2003. S. Müllegger and A. Winkler, Surface Sci. 600, 3982, 2006. J. Stettner, P. Frank, T. Griesser, G. Trimmel, R. Schennach, R. Resel, and A. Winkler, in this proceedings. P. Frank, J. Stettner, F. Nußbacher, and A. Winkler, in this proceedings. R. Resel, M. Oehzelt, T. Haber, G. Hlawacek, C. Teichert, S. Müllegger, and A. Winkler, J. Cryst. Growth 283, 397 2005. T. Haber, S. Müllegger, A. Winkler, and R. Resel, Phys. Rev. B74, 045419, 2006. P. Frank, G. Hlawacek, O. Lengyel, A. Satka, C. Teichert, R. Resel, and A. Winkler, Surface Sci. 601, 2152, 2007. P. Frank, G. Hernandez-Sosa, H. Sitter, and A. Winkler, Thin Solid Films 516, 2939, 2008. F. Balzer and H. G. Rubahn, Surface Sci. 548, 170, 2004. T. Haber, A. Andreev, A. Thierry, H. Sitter, M. Oehzelt, and R. Resel, J. Cryst. Growth 284, 209, 2005. S. Müllegger, G. Hlawacek, T. Haber, P. Frank, C. Teichert, R. Resel, and A. Winkler, Appl. Phys. A 87, 103, 2007. S. Müllegger, O. Stranik, E. Zojer, and A. Winkler, Appl. Surf. Sci. 221, 184, 2004.

Crystalline Stages of Rubrene Films Probed by Raman Spectroscopy B.A. Paez1, Sh. Abd-Al-Baqi2, G.H. Sosa2, A. Andreev1, C. Winder1, F. Padinger1, C. Simbrunner2 and H. Sitter 2 1

NANOIDENT Technologies AG, Untere Donaulände 21-25, A-4020 Linz, Austria E-mail: [email protected] 2 Institute of Semiconductor and Solid State Physics, University Linz, Altenbergerstr. 69, A-4040 Linz, Austria E-mail: [email protected] Abstract. We report on ex situ Raman characterization of rubrene thin films grown by hotwall epitaxy on cleaved mica substrates. Analysis of the vibrational bands revealed that at earliest growth stages the film is amorphous. In particular, a broad band at 1373 cm–1 proves the amorphous nature of the film. The rubrene molecules in amorphous phase are geometrically distorted, since the appearance of the Raman band at 1606 cm–1 is only infrared active for rubrene molecules with the C2h symmetry group. Further growth leads to seeding of spherulites in the amorphous matrix and further to their coalescence. Raman bands from isolated spherulites embedded in an amorphous matrix and from coalesced spherulites show polarization dependence (depolarization ratio < 0.6), thus demonstrating their crystalline nature. It is also found that the breathing mode (1003 cm–1) represents the rubrene fingerprint feature independent of layer crystallinity.

1.

Introduction

Understanding and engineering of molecular crystals is of great interest to achieve a tuned performance of organic-based devices. Rubrene, a tetraphenyl derivative of tetracene, has recently attracted much attention since hole mobili-ties as high as 20 cm2 V–1 s–1 for single crystals at room temperature have been reported. It is also found that the limited cofacial π-stack interactions result in rubrene in very efficient electronic coupling, which is consistent with the band regime at room temperature [1]. Raman spectroscopy plays a very important role in determining vibrational properties of the matter. The usefulness of the method is mainly due to its sensitivity and non-destructiveness, capability to provide information on chemical identity, charge states [2,3], processes at interfaces [4] and structural order [5]. Investigations of rubrene single crystals have been reported elsewhere [6]. Although similar structures as those reported by us have been used in organic field effect transistors [7], Raman investigations on thin films are still missing up to now.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_6, © Springer-Verlag Berlin Heidelberg 2009

37

38

B.A. Paez et al.

2.

Experimental Methods

The rubrene layers were deposited by hot-wall epitaxy [8] on cleaved mica substrates. Raman investigations were carried out ex situ in a Labram Aramis spectrometer from Jobin Yvon HORIBA. The structures were analyzed in the back scattering configuration and excited with the 784 nm (1.57 eV) light from a laser diode. Throughout this report the Porto coding for light polarization is adopted. It takes into account the propagation of the incoming and scattered beams together with the polarization, i.e., z(xy)z’: means the incident light is propagating along the z axis and then back scattered (z’); xy in the brackets labels the electric field polarizations with respect to a fixed reference plane holding the sample.

3.

Results and Discussion

3.1. Surface Morphology Three growth stages were selected for the vibrational investigations. The first one was homogenous rubrene layer (further called as amorphous matrix-it is explained in section 3.3), the second were isolated spherulites embedded in amorphous matrix, and finally coalesced spherulites forming a closed rubrene layer. Corresponding optical micrographs are shown in Figure 1. More details can be found in Ref. [9]. Fig. 1. Morphology of rubrene films in different growth stages: a) isolated spherulites in amorphous matrix; b) closed rubrene layer in coalescence stage. Raman measurements points: 1,2 – the center and close to border of an isolated or coalesced spherulite; 3 – amorphous matrix.

3.2. Vibrational Identity Rubrene films are first characterized by the breathing mode at 1003 cm–1. This mode proves the presence of the rubrene molecules on the mica substrate, i.e. it represents a fingerprint for their presence on a surface. The corresponding Raman spectra for the spherulites in the amorphous matrix and for the closed rubrene layer are shown in Figure 2.

Crystalline Stages of Rubrene Films Probed by Raman Spectroscopy 39 Fig. 2. Raman spectra (breathing mode) for isolated spherulites in amorphous matrix (top), and for the closed rubrene layer (bottom). Numbering as follows: (1)-spectra from the center, (2)-close to the border of the spherulite, (3)from amorphous matrix. Number 2 is also used for the closed layer, where all spherulites are coalesced. Rubrene powder and mica substrate spectra are also shown as reference.

Breathing (Ag) 1003 cm-1 λexc = 785 nm

50

z (xy) z ˙

cts, s-1

x3.6 x3.6 x2.4

z (yy) z ˙ z (xy) z ˙ z (yy) z ˙

x6.4

z (xu) z ˙

x6.6

z (xy) z ˙

x0.4 x1.0 x0.7 x1.0

z (yy) z ˙ z (xy) z ˙ z (yy) z ˙ z (xu) z ˙

50

x1.6 x1.0

Rubrene Powder Mica substrate

1000

2. 3. 1. 1. 2. 2.

cts, s-1

z (yy) z ˙

950

1. 1. 2.

Raman shift / cm-1

2. 2.

1050

The spectra were recorded at different laser light polarizations. In order to skip strong differences on the Raman intensity, all the spectra are normalized and the multiplication factors are indicated. It is found that the profile of the breathing mode is quite similar for amorphous matrix, isolated spherulites in amorphous matrix, and for closed rubrene layer, excepting the higher signal to noise ratio in the amorphous phase. 3.3. Amorphous Matrix and Crystallinity: Raman Investigations The use of polarized light in Raman spectroscopy allows to identify amorphous phases, crystallinity, isotropy, and in general to test selection rules of the investigated structure. The Raman spectra measured from the rubrene matrix material were found to be independent of light polarization and also the scattered light was left unpolarized (z(yu)z’). That means no preferred orientation exist in this part of the rubrene layer. Therefore we called it the amorphous matrix. In Figure 3, the Raman spectra of the spherulites in amorphous matrix and of the closed layer are shown. Although, the Raman spectrum of the amorphous phase lacks of several vibrational features like those observed in rubrene powder, there are clear signatures, which proves the presence of this phase in rubrene films. As it was already mentioned, the first one (but not exclusive) is the isotropy of the Raman spectrum upon light polarization. Second one is the broad band at 1373 cm–1 with a small shoulder at 1356 cm–1 (both labeled AM*). Namely, the both bands are clearly seen for amorphous phase, but completely missing for rubrene powder and closed rubrene layers. Isolated spherulites present intermediate

40

B.A. Paez et al.

case maybe due to the quite high penetration depth of the laser light in rubrene. Another characteristic of amorphous phase is the molecular symmetry breakdown proved by the broad band at 1606 cm–1. The mode is only infrared active according to the molecular symmetry group of isolated rubrene (C2h). The seeded in the amorphous matrix spherulites preserve most of the bands of rubrene powder (Figure 3). The Raman spectra, recorded both for the center and at the boundary of a spherulite, reveal strong polarization dependences, for example, for the bands depicted by the phenyl groups (1303 cm–1), and tetracene backbone (1522 cm–1). The symmetry of the Ag mode for the phenyl groups and Bg mode for the tetracene core turns from the coupling between the electric field and the dipole transition in the rubrene structure. AM˙ Phenyl groups 1303 cm-1 (Ag)1373 cm-1 λexc=785 nm

z (xy) z˙

Tetracene core AM˙ 1522 cm-1(Bg)1606 cm-1(Ag) 102 cts. s-1

X 8.0

z (yy) z˙

1.

z (xy) z˙

1. 2.

z (yy) z˙

Amorphous

z (xu) z˙

2. 3.

z (xy) z˙

102 cts. s-1

X 1.0

z (yy) z˙

1. 1.

z (xy) z˙ z (yy) z˙

2.

z (xy) z˙

2.

z (yy) z˙

2. 2.

1200

1300

1400

1500

Raman shift / cm-1

1600

Rubrene powder Mica substrate

Fig. 3. Raman spectra of rubrene films: spherulites in amorphous matrix (top), and from closed rubrene layer (bottom); 1,2 – the center and close to border of a spherulite; 3 – amorphous matrix. Rubrene powder and mica substrate spectra are also shown as reference. AM* - labels the signatures of the amorphous matrix.

The coalesced spherulites forming closed rubrene layer are also sensitive to the polarized light. Depolarization ratio between different polarized spectra (below 0.6) indicates the crystalline nature of the rubrene in this stage. A common feature of many spectra was the sloped background, despite of the fact that low excitation energy (1.57 eV) was used for measurements. It comes from the high photoluminescence yield of the rubrene [10]. In particular, the

Crystalline Stages of Rubrene Films Probed by Raman Spectroscopy 41

background of the spherulites is sensitive to the light polarization, which indicates selection rules for the luminescence (Fig. 3).

4.

Conclusions

Raman investigations proved that rubrene layers grown by hot wall epitaxy have different structural properties depending on the growth stage. Rubrene films independent on their amorphous or crystalline nature were distinguished by the breathing mode at 1003 cm–1 (Ag symmetry), i.e. this mode represents the rubrene fingerprint feature. Analysis of the vibrational bands revealed that at earliest growth stages the film is amorphous. In particular, a broad band at 1373 cm–1 proves the amorphous nature of the film. On the other hand, the mode at 1606 cm–1, usually an infrared active band, proves a symmetry breakdown of the molecule at this growth stage. Additionally, the amorphous phase lacks of vibrational activity at the phenyl groups, and tetracene backbone. Therefore, it is likely that the geometry of the rubrene molecule is dramatically distorted. Further growth leads to seeding of spherulites in the amorphous matrix and further to their coalescence. This growth phase associated with spherulite-like shapes (spherulites in amorphous matrix and coalesced spherulites) was found to be highly sensitive to the light polarization, which is shown by the phenyl group band (1303 cm–1) and the tetracene core band (1522 cm–1). That proves clearly the crystalline nature of the rubrene in the spherulites.

References 1

D. Beljonne, et al., in “Handbook of conducting polymers”, Conjugated polymers: theory, synthesis properties and characterization, chap. 1, 3rd ed. Edited by T. A. Skotheim and J. R. Reynolds (CRC Press, 2007). 2 B. A. Paez, et al., Proc. SPIE Int. Soc. Opt. Eng. 5217, 63 (2003). 3 M. L. Shand, W. Richter, E. Burstein, and J. G. Gay, J. Nonmmetals 1, 53–62 (1972). 4 B. A. Paez, et al., Appl. Surf. Sci. 234, 168 (2004). 5 P. Colomban, Spectroscopy Europe 15 (6), 8 (2003). 6 J. R. W.-Wolf, L. E. McNeil, S. Liu and C. Kloc, J. Phys.: Condens. Matter 19 (2007) 276204 (15 pp). 7 S.-W. Park, et al., Appl. Phys. Lett. 90, 153512 (2007). 8 H. Sitter, chap. 5, in Organic Nanostructures for Next Generation Devices. Edited by K. Al-Shamery, H.-G. Rubahn, and H. Sitter (Springer-Verlag Berlin Heidelberg 2008). 9 Sh. Abd al-Baqi, et al., Proceedings Symposium O, E-MRS Spring Meeting 2008, Springer “Proceedings in Physics”, in print. 10 S. 8, Z. Peng, X. Zhang, S. Wu, Journal of Luminescence 121, 568–572 (2006).

Rubrene Thin Film Characteristics on Mica Sh. M. Abd Al-Baqi1, G. Henandez-Sosa1, H. Sitter1, B. Th. Singh 2, Ph. Stadler 2 and N. S. Sariciftci 2 1

Institute of Semiconductor and Solid State Physics, Johannes Kepler University, Linz, Austria 2 Institute of Physical Chemistry and Linz Institute For Organic Solar Cells (LIOS), Johannes Kepler University, Linz, Austria E-mail: [email protected]

Abstract. Rubrene thin films were deposited by Hot Wall Epitaxy on mica substrates.

To optimise the growth conditions, the growth rate and the substrate temperature were changed systematically. The surface morphology of the grown rubrene layers was investigated by polarized optical microscopy (POM), electron microscopy (SEM) and atomic force microscopy (AFM). After an initial nucleation and coalescence stage a continuous amorphous layer is formed. In a later stage of growth, spherulites embedded in the amorphous matrix are found, which furthermore cover the whole surface. It could be proven that the spherulite consist of polycrystalline material, which could be used for the fabrication of organic field effect transistors.

1.

Introduction

Many attempts were made to fabricate organic field effect transistors from rubrene thin films [1–4]. The largest mobility obtained so far in rubrene OFETs is 2.5 cm2. V–1. s–1, which is still much less than in monocrystlline bulk material [5]. The main difference between bulk and thin film material is the crystalline property. Consequently, the main effort goes in the direction of improving the crystallographic order in the rubrene layers. Since there is no lattice matched substrate available, the only chance to approach the goal of highly ordered structures is to use an optimised growth regime. Due to the weak Van der Wales type bonds of the molecules, a deposition process as close as possible to thermodynamic equilibrium seams to be appropriate. The local dynamic equilibrium at the growing surface allows a manifold impingement and re-evaporation of the molecules. Supported by an enhanced surface mobility the rubrene molecules can find the optimum position on the substrate to form a highly ordered structure, which can lead to a layer of enhanced crystallinity. The method of choice to provide such growth conditions is the Hot Wall Epitaxy (HWE).

2.

Experimental Procedure

The rubrene source material was purchased from Aldrich (purity > 98%) and purified by threefold sublimation under dynamic vacuum conditions. The rubrene layers were evaporated in a standard HWE reactor on freshly cleaved 2M1 muscovite mica. Two different growth rates were used by employing different

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_7, © Springer-Verlag Berlin Heidelberg 2009

43

44

Sh.M.A. Abd Al-Baqi et al.

source temperatures (Ts) for the evaporation of rubrene (Ts = 180°C and 235°C). The surface mobility of the rubrene molecules was influenced by the substrate temperatures (Tsub = 80°C, 90°C or 120°C). The surface structure of the rubrene layers was investigated routinely by polarized optical microscopy. Atomic Force Microscopy (AFM), (VEECO Di3100) in tapping mode was employed. In the case of high surface roughness SEM was used to investigate the surface morphology. The crystalline property of the rubrene layers was probed by polarization depend Raman spectroscopy and x-ray diffraction experiments reported in detail elsewhere [6,7].

3.

Results and Discussion

Since mica substrates can be freshly cleaved before evaporation and therefore provide a very clean surface, we selected this substrates to investigate the growth process of rubrene layers. The cleaving process was done ambient air, which means that some dust particles can still contaminate the substrate surface. In the early stage of growth single rubrene islands are formed which coalesce and form an amorphous layer [8]. The later stage of growth is dominated by the formation of spherulites. Figure 1 summarizes the optical micrographs to demonstrate the surface structure after different growth times. Very similar rubrene surface structures were observed previously on SiO2 substrates [9]. The development of spontaneously nucleated spherulites is clearly seen, which are embedded in an amorphous matrix. By increasing the deposition time, the spherulites grow in diameter and consist of an inner disc surrounded by an outer ring. Finally the spherulites coalesce and cover with the outer ring structures the whole substrate. Depending on their growth mechanisms spherulites can be divided into two categories nucleated either from single nucleation site or bunch of needles and both categories have been found in these thin films from TEM measurement [10]. The radii of the central disc and the outer rings are plotted as a function of growth time in Fig. 2, showing a linear increase with the onset of a saturation due to coalescence of the spherulites. The nucleation of the spherulites can be caused by static impurities or dynamic heterogeneities [10]. If the pin holes in the amorphous matrix would act as nucleation centers, as assumed in ref. [9], the density of spherulites should be the same as the density of the pinholes, which is by far not the case. On the other hand, if the spherulites are originated by defects in the amorphous larger, the spherulites would grow on top of the amorphous matrix which is also not the case. As shown by the AFM picture in Fig. 3, the spherulites are embedded in the amorphous matrix. The cross section across the border of the spherulite (see Fig. 3) shows the same height for the spherulites as for the amorphous matrix and a clear trench at the borderline. So we assume that starting from an impurity the amorphous matrix undergoes a phase change by forming poly crystalline structures. In that way the material of the amorphous matrix is consumed until the whole substrate is covered with spherulites.

Rubrene Thin Film Characteristics on Mica

45

Rubrene/Mica

Ts = 235°C Twall = 235°C Tsub = 80°C Growth time = 30 min Growth time = 1 hour Growth time = 3 hours

2 0 0 µm

5 0 0 µm

2 0 0 µm

Fig. 1. Optical Micrographs of the surface structure with increasing growth time. 1000 900

Radius (µm)

800 700 600 500 400 300 200

Inner Disc Outer Disc

100 0 0

100

200

300

t (minutes)

Fig. 2. Diameter of the inner disc and outer rings as a function of growth time, Ts = 235°C, Twall = 235°C, Tsub = 90°C.

trench

spherulite

amorphous

10µm

Fig. 3. AFM image of rubrene on mica together with a cross section across the border.

Optical microscopy using polarized light gives a first hint on the crystallinity of the obtained structures. A typical result of optical micrographs obtained without polarization, parallel and perpendicular orientation of the polarizer and analyzer are shown in the Fig. 4. If the relative position of the polarizer and analyser is changed; the different regions of the spherulites change their colour, while the

46

Sh.M.A. Abd Al-Baqi et al.

surrounding amorphous matrix stays unchanged. This can be interpreted, that the spherulites consist of small crystallites oriented in a radial direction. A detailed investigation of the crystallographic order inside the spherulite was performed by polarization dependent micro Raman spectroscopy and x-ray diffraction [6,7]. Due to the higher surface roughness inside the spherulite the morphological details were studied by SEM. Figure 5a shows the typical structure of the inner disc of a spherulite. Zooming into the central part two different features can be observed. Out of a layer consisting of similar elongated mosaic blocks, (Fig. 5b) whisker like facetted structures grow in the third dimension (Fig. 5c). During the growth of the spherulites in lateral direction by recrystallization of the amorphous material, the flux of impinging rubrene molecules continuous. The additional molecules hitting the amorphous matrix contribute to the continuous growth of this part of the layer. The other molecules impinging on the polycrystalline spherulites find crystalline mosaic blocks as seeds for the formation of whiskers, pointing out of the surface in the central part of the sphurlites. Figure 4c clearly shows the faceted structure of such whiskers. Rubrene/Mica Tsource = 235°C Twall = 235°C Tsub = 80°C growth time = 1 hour Without polariztion 180° 90°

200µm

200µm

200µm

Fig. 4. Rubrene thin film on mica substrate under polarized optical microscope.

200 nm

1 μm

200 nm

Fig. 5. SEM pictures for rubrene on mica substrate Ts = 230°C, Twall = 240°C, Tsub = 120°C, Tpre-heat = 90°C.

4.

Conclusions

Rubrene thin films were evaporated by HWE on mica. After the formation of an amorphous matrix, spherulite structures start to grow, which finally cover the whole surface. Due to their polycrystalline property the spherulites are resistant against oxidization while the amorphous matrix becomes transparent upon exposure to air, which is a clear sign for oxidation. The rubrene spherulites provide therefore a promising material for the fabrication of OFETs.

Rubrene Thin Film Characteristics on Mica

47

Acknowledgements. The work was supported by the Austrian Science Foundation (FWF) within the National Research Network (NFN) “Interface controlled and Functionalized Organic Films”.

References 1 2 3 4 5 6 7 8 9 10

Se-W. Park, et al., Appl. Phys. Lett. 90, 153512 (2007). Se-W. Park, et al., Appl. Phys. Lett. 91, 033506 (2007). M. Nothaft et al., Phys. stat. sol. (b) 245, 788–792, (2008). C.H. Hsu, et al., Appl. Phys. Lett. 91, 193505 (2007). V. Podzorov, et al., Phys. Rev. Lett. 93, 086602 , (2004). T. Djuric, et al., E-MRS proceedings 2008, (in print). B.A. Paez, et al., E-MRS proceedings 2008, (in print). Gregor Hlawacek, et al., E-MRS proceedings 2008, (in print). Y. Luo, et al., phys. Stat. Sol. (a) 204, No. 6, 1851–1855 (2007). L. Gránásy, et al., Physical Review E 72, 011605 (2005).

Structural Properties of Rubrene Thin Films Grown on Mica Surfaces T. Djuric1, H.-G. Flesch1, M. Koini1, Sh.M. Abd Al-Baqi2, H. Sitter2, and R. Resel1 1 2

Institute of Solid States Physics, Graz University of Technology, Austria Institute of Semiconductor and Solid States Physics, University Linz, Austria E-mail: [email protected]

Abstract. Structural properties of rubrene thin films on cleaved mica (001) surfaces were investigated by optical microscopy and x-ray diffraction. Optical microscopy shows, that the crystallization of rubrene results in formation of spherulites. X-ray specular diffraction reveals polycrystalline and polymorphic nature of rubrene. The pole figure measurements of films prepared at low deposition rates reveal orthorhombic structure and indicate fiber textures with crystallographic planes (121), (131) and (141) preferentially oriented parallel to the substrate surface. High deposition rate thin films in addition show polymorphism, corroborating the existence of the orthorhombic and the triclinic phase.

1.

Introduction

Rubrene is an aromatic organic molecule with many desirable properties, which make it to a material of choice for the fabrication of single crystal organic field effect transistors. It has the highest reported charge carrier mobility in organic single crystal transistors (20 cm2/Vs) [1]. It is proposed that pristine crystalline rubrene is rather insensitive to oxidation due to its packing [2]. In spite of its very promising electronic behavior as a single crystal, the fabrication of rubrene thin films turned out to be a difficult task. Many efforts to grow well structured rubrene thin films on different substrates resulted in amorphous films with only small, usually spherulitic shaped, crystalline areas [3–5]. This growth of rubrene thin films can be divided into two subsequent stages. In the initial growth stage amorphous islands are formed [4,5]. When a critical coverage is obtained, amorphous islands merge together, coalescence starts and amorphous porous thin films are obtained [5]. The porous film serves as a template for the second growth stage and enables the nucleation of polycrystalline spherulites [5]. Measured charge carrier mobility of crystalline spherulitic areas revealed very small values, ranging between 10–6 and 10–3 cm2/Vs. Rubrene thin films reported here evinced the same morphological properties. Rubrene crystallizes in three different polymorphic phases. The first reported crystal structure of the rubrene is monoclinic (a = 15.500Å, b = 10.100Å, c = 8.800Å, β = 90.55°) [6]. Also a triclinic structure is described (a = 9.150Å, b = 11.600Å, c = 7.160Å, α = 103.53°, β = 112.97°, γ = 90.98°) [7]. Recent investigations report a orthorhombic structure (a = 7.184Å, b = 14.433Å and c = 26.897Å), which is the only one where a full structure solution is available [8–10]. H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_8, © Springer-Verlag Berlin Heidelberg 2009

49

50

T. Djuric et al.

2.

Experimental Methods

Rubrene was purchased from Aldrich (elemental purity > 98%) and additionally purified by gradient sublimation. Freshly cleaved mica (001) was used as substrate. Rubrene thin films were deposited by hot wall epitaxy in a vacuum chamber with a base pressure below 10–4 Pa at different deposition rates and substrate temperatures (Ts). Pole figures were measured with a Philips X’pert x-ray diffractometer using CrKα radiation and a secondary side graphite monochromator. Specular scans were performed on a Bruker D8-Discover diffractometer using CuKα radiation. POWDER CELL and STEREOPOLE were used for the evaluation of the specular scans and simulation of pole figures.

3.

Results and Discussion

Figure 1. (a)–(d) shows optical micrographs of the rubrene thin films, which were deposited under different growth conditions. Samples (a)–(c) were grown for 24h at low deposition rate (LDR) (Tsource = 180°C, Twall = 180°C) but at different Ts: 80°C, 90°C and 120°C. Sample (d) was grown for 3h at a high deposition rate (HDR) (Tsource = 235°C, Twall =235°C) while the Ts was held at 120°C. All samples exhibit spherulitic morphologies. Spherulites are aggregates of microcrystals, which arrange in radially growing fiber-like structures. At low Ts and low deposition

Fig. 1. Optical microscopy images of rubrene films deposited on mica substrates. Samples (a)–(c) were grown at low deposition rate at substrate temperatures: (a) 80°C, (b) 90°C and (c) 120°C. Sample (d) is grown at a high deposition rate. Micrographs below present growth stages of rubrene spherulites. Nucleating as single needle bunches (α), splaying through small-angle branching (β) to a spherulite with a spherical envelope (γ).

Structural Properties of Rubrene Thin Films Grown on Mica Surfaces

51

rubrene orth (020)

rubrene rubrene

rubrene

rubrene

rubrene

mica(006)

rubrene orth(002)

(1) LDR T sub= 80° (2) LDR T sub= 90° (3) LDR T sub= 120° (4) HDR T sub= 120°

mica(004)

log(intensity)

rubrene tric (001)

mica(002)

rates (LDR) thin films consist of separate spherulites. With increasing substrate temperature or with higher deposition rate spherulites are growing until impingement. In Fig. 1(c) one can see rather straight boundaries of impinging spherulites This indicates that neighboring spherulites start to grow at the same time with the same growth rate [11]. Otherwise hyperbolic boundaries are obtained, which partially appear in Fig. 1(d) (marked with arrows). Please note that the straight boundaries between the spherulites could appear due to cleavage steps of mica. Depending on their growth behavior spherulites can be divided in two categories [12]. Category I shows radial growth starting from a single nucleation site. Category II spherulites nucleate as single bunches of needles, through small-angle branching the needles spread until a spherical envelope is formed. Figure 2 (α)-(γ) shows the formation of spherulites observed on films grown with a HDR for 15min which indicates spherulites of category II. Figure 2 shows specular scans of rubrene thin films, measured with a specular offset of 3°. The position of the Bragg peaks enables an identification of the crystalline structure together with a preferred orientation of the crystallites. Since the thin film volume is mainly composed of spherulites, it can be supposed that the measured diffraction intensities reflect the crystallographic properties of the spherulites. For the phase identification primarily peaks at low values of the scattering vector qz were used. In comparison with the calculated values measured peak positions at qz = 0.47Å–1 and qz = 0.87Å–1 can be assigned unambiguously to net planes (002) and (020) of the orthorhombic phase, while the measured position at qz = 0.55Å–1 is in a good agreement with the calculated net plane (010) of the triclinic phase. Six other rubrene peaks were detected, which cannot be assigned unambiguously; even the monoclinic phase cannot be excluded definitely. Measured diffraction patterns of the LDR samples don’t show significant diffraction peaks, although crystalline rubrene has to be present.

rubrene

(4) (3) (2) (1)

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

-1

q z [Å ]

Fig. 2. Specular scans of rubrene films grown on mica. Weak diffraction intensities are observed for the low deposition rate (LDR). The orthorhombic phase with (002), (020) and the triclinc phase with (010) parallel to the substrate can be assigned for the high deposition rate (HDR) film. Curves are shifted for clarity.

52

T. Djuric et al.

In Figure 3 characteristic pole figures of LDR and HDR films are shown. In case of the LDR film (Fig. 3a,b) a so-called fiber texture is observed. The rather blurred features in Ψ-direction can be explained by multiple crystal orientations which are slightly tilted against each other. Due to defocusing effects the position of pole densities are shifted to smaller values. The crystallographic net planes (121), (131) and (141) of the orthorhombic structure are found to be parallel to the mica surface. For these orientations the aromatic backbone of the molecule is nearly parallel to the substrate surface. These planes were not observed in the specular scan, because their positions overlap with Bragg peaks of mica. In Fig. 3 c,d pole figures of the HDR film, which also reveal a fiber texture, are shown. By the reason of the huge variety of crystal orientations parallel to the mica surface on the one side and the existence of polymorphic phases on the other side, multiple orientations are necessary to explain the pole figures. One additional difficulty is given by the fact, that a full structure solution for the triclinic phase is not known; hence no prediction about high Bragg intensities can be given. Simulated positions of enhanced pole densities with net planes (001) and (010) of the orthorhombic and (010) of the triclinic phase parallel to the substrate surface are in good agreement with the measured pole figures. These orientations are corroborated by the specular scan (Fig. 2).

Fig. 3. Pole figures of rubrene films prepared at low deposition rate are measured at qz = 0.86Å–1 (a) and qz = 1.19Å–1 (b). The calculated pole densities of individual crystal orientations with (141), (131) and (121) are denoted by , and , respectively. High deposition rate films are measured at qz = 1.45Å–1 (c) and qz = 1.72Å–1 (d). The calculated pole densities due to (001), (010) of the orthorhombic phase and (010) of the triclinic phase are denoted by , and , respectively. The single high intensity spots are due to single crystalline mica substrate.

Structural Properties of Rubrene Thin Films Grown on Mica Surfaces

4.

53

Conclusions

Spherulitic growth of rubrene thin films grown on mica (001) surfaces is observed. The growth behavior of rubrene spherulites is assigned to the category II, which nucleates from single bunches of needles. By means of x-ray specular scans and pole figure measurements it is found that spherulites are polycrystalline and polymorphic crystal entities. Pole figures of the LDR film indicate a widely smeared fiber texture with net planes (121), (131) and (141) of the orthorhombic structure oriented parallel to the substrate surface. For these orientations the aromatic backbone of the molecule is nearly parallel to the substrate surface and consequently an identical orientation of the molecules within the first monolayer of spherulites is proposed. Pole figures of the HDR thin film also reveal a fiber texture but with other preferred orientations. Here orthorhombic net planes (001), (010) and the triclinic net plane (010) are parallel to the mica surface. Orthorhombic net planes (001) and (010) correspond to cleavage planes in rubrene. Acknowledgements. The work was supported by the Austrian Science Foundation (FWF) within the National Research Network (NFN) “Interface Controlled and Functionalized Organic Films”.

References 1

V. Podzorov, E. Menard, A. Borissov, V. Kiryukhin, J.A. Rogers and M.E. Gershenson, in Phys. Rev. Lett., Vol. 93, 8, 2004. 2 O. Mitrofanov, C. Kloc, T. Siegrist, D.V. Lang and W.Y. So, A.P. Ramirez, in Appl. Phys. Lett., Vol. 91, 212106, 2007. 3 F. Balzer, M. Schiek, A. Lützeu and K. Al-Shamery in Proc. SPIE, 64706, 2007. 4 Y. Luo, M. Brun, P. Rannou and B. Grevin in Phys. stat. sol., Vol. 204, 1851, 2007. 5 S. Seo, B. Park and Paul G. Evans, in Appl. Phys. Lett., Vol. 88, 232114, 2006. 6 W.H. Taylor in Z. Kristall., Vol. 93, 151, 1931. 7 S.A. Akopyan, R.L. Avoyan and Yu.T. Struchkov in Zh. Strukt. Khim, Vol. 3, 602, 1962. 8 D.E. Henn, W.G. Williams and D.J. Gibbons in J. Appl. Cryst., Vol. 4, 1971. 9 I. Bulgarovskaya, V. Vozzhennikov, S. Aleksandrov and V. Belsky in Latv. PSR Zinat. Akad. Vestis Fiz. Teh. Zinat. Ser., Vol. 4, 53, 1983. 10 I.D. Jurchescu, A. Meetsma and T.T.M. Palstra in Acta Cryst. B, Vol. 62, 330, 2006. 11 H. Müller, in phys. stat. sol. (a), Vol. 66, 199, 1981. 12 L. Granasy, T. Pusztai, G. Tegze, J.A. Warren and J.F.Douglas in Phys. Rev. E, Vol. 72, 011650, 2005.

Rubrene on Mica: From the Early Growth Stage to Late Crystallization Gregor Hlawacek1, Shaima Abd-al Baqi2, Xiao Ming He1, Helmut Sitter2 and Christian Teichert1 1

Institute of Physics, University of Leoben, Leoben, Austria E-mail: [email protected], [email protected] 2 Institute of Semiconductor and Solid State Physics, University of Linz, Linz, Austria Abstract. The fabrication of Rubrene thin films is of interest because of the high mobility observed for Rubrene single crystals. Here, we report on an atomic force microscopy (AFM) investigation of the growth of Rubrene thin films by Hot Wall Epitaxy on mica(001). During the initial formation of amorphous islands, a non-constant growth rate is observed due to temperature dependent changes in the sticking coefficient. Furthermore, the contact angle of these islands – also measured by AFM – depends on temperature. With continuous deposition, island coalescence starts resulting in ramified surface aggregates. The final growth stage is characterized by the formation of crystalline spherulites which also analyzed by AFM.

1.

Introduction

Rubrene is known for its large carrier mobility in single crystal form. Values of up to 20 cm²/Vs have been reported [1,2]. Although this would make it an interesting candidate for various devices, thin films grown from Rubrene have shown hole mobilities lower by 7 orders of magnitude [3]. Unfortunately, very often these thin films are amorphous. Recent studies have shown that in crystalline or polycrystalline films grown with a large overpressure of Rubrene or by utilizing heterostructures, mobilities of up to 0.2 cm²/Vs can be reached [4–6]. Recent work using the weakly interacting substrate SiO2 and an OTS layer to tune the surface energy of the substrate demonstrated mobilities between 0.1 cm²/Vs and 2.5 cm²/Vs [7]. Here, thin films of Rubrene are grown on mica(001) by means of hot wall epitaxy (HWE). The initially formed amorphous islands as well as the morphologies in crystalline spherulites, observed in thick films, are characterized by atomic force microscopy (AFM). The behavior of the growth rate, the contact angle of the Rubrene islands, and their fractal dimension are analyzed in dependence on growth temperature and film thickness.

2.

Experimental Methods

Mica(001). The (001) surface used is a cleavage plane of 2M1 muscovite. It allows easy ex situ cleavage offering large (>100 µm) atomically flat terraces. Furthermore it has shown the capability to align organic molecules by the a strong surface dipole [8].

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_9, © Springer-Verlag Berlin Heidelberg 2009

55

56

G. Hlawacek et al.

Standard quality mica was cleaved before insertion into the HV chamber, thus a fresh cleaved surface with only a few cleavage steps is used. Hot Wall Epitaxy (HWE). HWE is a high vacuum variant of physical vapor deposition with a base pressure of 10–6 mbar [9]. In contrast to many other growth techniques it utilizes the near field of the molecular beam by moving the sample close or even into the hot wall tube that holds the film material. The walls of the tube can be heated separately and are held on a higher temperature than the sample and the source. This prevents deposition on the tube wall and helps to create a uniform flux of molecules. The main advantages of HWE are that the films are grown close to the thermodynamic equilibrium. The main drawback is that the position of the sample close to the evaporator makes an in situ characterization of the film growth impossible. Rubrene. This organic semiconductor consists of a tetracene core with four additional phenyl groups connected by single bonds. In contrast to the desired crystalline phase the amorphous one is not stable against oxidation [10]. The two states can be distinguished easily, since the amorphous phase lacks the typical red color found for crystalline films. Atomic Force Microscopy (AFM). We used an Digital Instruments MultiMode IIIa AFM in intermittant mode to avoid damage to the organic thin film. Conventional Si probes with opening angles of 20° and tip radii of less then 10 nm have been employed. The typical resonance frequency of the used cantilevers is 300 kHz and the force constant is about 40 N/m.

3.

Results and Discussion

Figure 1 shows two series of AFM images obtained from amorphous Rubrene (or more likely oxidized Rubrene [10,11]) thin films grown with a substrate 363 K (a–d) and 393 K (e–h). The deposition times for both image sequences ranged between 2 min and 24 h. First, small circular islands of rather uniform size are formed which then grow (a,b,e,f) and start to coalesce (c,g,h). In Fig. 2 the distribution of island height and island base area for the case of 363 K is presented. The island height changes from 50 nm after 2 min to 120 nm after 60 min. While the width of the height distribution remains constant (20 nm) with ongoing deposition, the width of the island area distribution, however, broadens dramatically as soon as coalescence starts. The same holds for the films grown at higher temperature (lower row of Fig. 1). However, in this case the island density is much lower which can be related to the increased mobility of the molecules on the surface at higher temperatures. Figure 3(a) shows the nominal film thickness ( f ) vs. deposition time. These thickness values are obtained by calculating the total volume of Rubrene divided by the image area of at least three independent images. For the sample grown at higher temperature, f ranges significantly below the film thickness obtained under identical conditions at lower temperature. We can explain this by a change in sticking

Rubrene on Mica

57

coefficient for Rubrene on mica(001). From the power law fits indicated we obtain exponents of nearly unity (linear dependence as expected) for the higher temperature growth and 1/2 for the low temperature growth. We can understand this behavior if we assume different sticking coefficients, namely a high one for Rubrene on mica(001) and a low one for Rubrene on Rubrene. The observed change in lateral shape of the islands can be evaluated by the fractal dimension D. Figure 3(b) shows the change in D vs. deposition time, where D is calculated by applying a linear fit to the power spectrum [12] of the corresponding 10 µm ×10 µm AFM images. Again, the growth of the amorphous Rubrene islands can be divided in three stages. First, the compact islands are formed and the highest fractal dimension is obtained. Then coalescence starts, and the fractal dimension is reduced as more one-dimensional aggregates are formed. The last stage shown in Fig. 1(d) leads to a further reduction of the dimensionality as the islands become more ramified.

Fig. 1. AFM images of Rubrene thin films grown by HWE on mica(001). Top row samples grown at 363 K, bottom row sample temperature 393 K. Deposition times are (a,e) 2min, (b,f) 15 min, (c,g) 60 min, and (d,h) 24 h.

Fig. 2. Island height (left) and island size histograms for Rubrene films grown at 363 K.

58

G. Hlawacek et al.

From the islands three dimensional shape we can make qualitative estimations on the surface free energy. As the Rubrene islands in the initial stage have the shape of a droplet it is assumed that they are amorphous which is confirmed by several techniques [3,13]. Thus, AFM allows measuring the contact angle Ɵ, like for liquids, by analyzing cross sections through the island center [3]. Figure 4 shows two cross sections through islands grown at different temperatures. Both sections are taken from samples with deposition times shorter than the deposition time required for coalescence to allow the largest possible drop volume but before changes in island shape might influence the contact angle. Besides the obvious size difference, the contact angle for the film grown at elevated temperature is larger than the one grown at lower temperature. This is contrary to what would be expected at first glance, since surface energy in general decreases with increasing temperature. The observed behavior can be explained when considering different temperature dependencies of surface energies for Rubrene and mica(001). It seems that the surface energy of Rubrene decreases slower with increasing temperature as the one of mica. As a result the surface becomes more rubrenophobic with increasing temperature.

Fig. 3. (a) Nominal film thickness f from AFM images vs. deposition time. (b) Fractal dimension of the Rubrene islands vs. deposition time. Dashed lines separate the regimes.

Fig. 4. AFM cross section through Rubrene islands. The contact angle is 27° for 393 K and 22° for 363 K.

The later growth stage is characterized by the formation of large spherulites [14]. Figure 5 shows AFM images obtained from different parts of a spherulite typically found after 1 hour of Rubrene deposition with a source temperature of 508 K and a sample temperature of 363 K. Three areas can be distinguished optically: a dark red center region, a lighter iris region, and a transparent matrix. Already from the red color impression we conclude that the spherulites are crystalline. The matrix, however, has turned transparent when the sample was removed from the high vacuum system because the amorphous Rubrene film is not stable against oxidation. The center (Fig. 5(a)) is formed by a rough (rms-roughness: 33 nm), highly crystalline and faceted area that exhibits a slight radial orientation. The surrounding

Rubrene on Mica

59

iris area (Fig. 5(b,c) rms-roughness 10 nm) shows branched, strongly radial orientated structures, typical for spherulitic growth. The edge of the spherulite is separated from the amorphous matrix by a deep trench (Fig. 5(c)). Furthermore, there is a significant change in height between the spherulite and the matrix region. Both observations are indications of a large mass transport towards the spherulite since it is fed from the amorphous matrix. Figure 5(d) finally represents the amorphous matrix (rms-roughness 0.7 nm) which is covered by many small holes of a few nanometer depth. These holes could be either due to the mentioned massive mass transport but are more likely a result of the oxidation process, since they exit all over the surface.

Fig. 5. Details of a Rubrene spherulite. (a) Center, (b) Iris, (c) spherulite rim, and (d) surrounding amorphous matrix. Z-scale: (a) 300 nm, (b) 50 nm, (c) 30 nm and (d) 5 nm. The insets show the cantilever position relative to the spherulite center.

4.

Conclusions

Quantitative morphological AFM analysis of HWE growth of Rubrene on mica(001) allowed to draw the following conclusions with respect to sticking coefficient and surface energies: Material and temperature depended sticking coefficients lead to non-linear growth rates for amorphous Rubrene on mica. The sensitivity of Rubrene growth with respect to the growth temperature is also reflected in the observed increase of the contact angle for Rubrene on mica(001) with increasing deposition temperature. Acknowledgments. Funding by Austrian Science Fund Projects S9707, S9706.

References 1 2 3 4 5 6 7

(a) A. L. Briseno, et al., Adv. Mater. Vol. 18, 2320, 2006; (b) V. Podzorov, et al., Phys. Rev. Lett. Vol. 93, 086602, 2004. M. E. Gershenson, et al., Rev. Mod. Phys. Vol. 78, 973, 2006. S. Seo, et al., Appl. Phys. Lett. Vol. 88, 232114, 2006. D. Kafer and G. Witte, Phys. Chem. Chem. Phys. Vol. 7, 2850, 2005. J. H. Seo, et al., Appl. Phys. Lett. Vol. 89, 163505, 2006. S.-W. Park, et al., Appl. Phys. Lett. Vol. 91, 033506, 2007. C.-H. Hsu, et al., Appl. Phys. Lett. Vol. 91, 193505, 2007.

60 8 9 10 11 12 13 14

G. Hlawacek et al. (a) H. Plank, et al., Thin Solid Films, Vol. 443, 108–114, 2003, (b) F. Balzer, et al., Appl. Phys. Lett., Vol. 79, 3860–3862, 2001. M. A. Herman, H. Sitter, Molecular Beam Epitaxy, Springer, 1989. M. Kytka, A. Gerlach, J. Kováč, F. Schreiber; Appl. Phys. Lett. 90, 131911 (2007). A. Otomo, et al., Opt. Let., Vol. 27, 891–893, 2002. A. Mannelquist, et al., Appl. Phys. A Vol. 66, 891, 1998. T. Djuric et al., this proceedings volume. (a) Y. Luo, et al., phys. stat. sol. (a) Vol. 204, 1851, 2007; (b) S. Abd-al Baqi et al., this proceedings volume.

β-Sheeted Amyloid Fibril Based Structures for Hybrid Nanoobjects on Solid Surfaces V. Bukauskas1, V. Strazdienė1, A. Šetkus1, S. Bružytė2, V. Časaitė2, and R. Meškys2 1 2

Semiconductor Physics Institute, Gostauto 11, Vilnius, Lithuania Institute of Biochemistry, Mokslininku 12, Vilnius, Lithuania E-mail: [email protected]

Abstract. A self assemblage of three-dimensional β-sheet protein-based supramolecular structures on solid state surfaces are investigated. A set of hybrid proteins containing the abeta40 peptide domain is constructed. Dimeric glucose dehydrogenases and thioredoxin are fused to abeta40 peptide. The supramolecular structures are immobilized on solid surfaces and the properties of the surface nanobjects are studied. Based on analysis of morphology and mechanical properties of these objects it is proved that Aβ40 peptide containing proteins preferably self assemble into island and grain type structures diameter of which is about 20–120 nm. In contrast to this TrxAβ40 fusion proteins preferably form thick (about 7–14 nm) and short (about 1–2 μm) objects. It is experimentally demonstrated that arrangement of fibrils on solid surfaces can be varied by duration of immobilization period and material of solid substrate.

1.

Introduction

During the last decade, artificial smart systems include large number of highly integrated single elements and completed modules dimensions of which extremely decreases due to advantages of nanotechnology. A progress in nanotechnology is partly achieved due to studies of self-assembling of supramolecules and hybrid structures and adopting of so called “bottom-up” approaches. It is extremely increasing interest in integration of biomolecules with nanosized materials aiming to create hybrid materials that combine the evolutionary optimized recognition and catalytic properties of biomaterials with the unique electronic, optical and catalytic functions of nanomaterials. Notable success of these emerging technologies in practical applications stimulates frontier research and development activities aimed at creation of new nanomaterials and nanostructures that are based on self-assemblage of supramolecules [1,2]. It is known that many proteins and peptides aggregate into extended β sheet like structures [3]. Formation of amyloid fibrils is recognized important not only in understanding of human diseases but also is found acceptable for development of ordered nanostructures with modifiable properties and applicable in biotechnology, material science, molecular electronics and related fields. Present study deals with specific aspects of development of hybrid proteins capable to create β-sheeted protein fibrils and integration of the fibrils with solid surfaces. Our study aims to define the most essential conditions and parameters on which the arrangement of the complex object on the solids depends. Morphology,

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_10, © Springer-Verlag Berlin Heidelberg 2009

61

62

V. Bukauskas et al.

mechanical and electrical properties are investigated by scanning probe microscopy in self-assembled complex hybrid systems on solid surfaces.

2.

Samples and Experimental Methods

A set of hybrid proteins containing the Aβ40 peptide domain and dimeric glucose dehydrogenase (GDH) or thioredoxin (Trx) have been constructed in present work. Expression and purification of the proteins was accomplished using prokaryotic system based on Escherichia coli. Details of the construction and expression of hybrid proteins are subject of special report and will be presented elsewhere. Hen egg lysozyme was fibrilized as described in our recent work. Photoluminescence spectra were measured for the hybrid proteins in a solution in the interval of wavelengths between 460 and 560 nm. The wavelength of stimulating light was 446 nm. Binding of thioflavin T to the proteins resulted in distinctive increase of the intensity of the photoluminescence. The supramolecular structures were immobilized on solid surface by deposition of biomolecular objects from colloidal solution. Insulating SiO2 layer on Si, SnO and In2O3 thin films and mica sheets were used for the substrates of combined structures. The substrate was placed on the surface of the liquid and pulled off the liquid after some fixed period of time between 1 and 30 minutes. The samples were dried in air with relative humidity about 30% and surrounding temperature about 295 K. The drying time was about 3 hours. Structure of the sample surfaces and electrical properties were investigated by SPM D3100/Nanoscope IVa (Veeco, Digital Instruments). The surface properties were characterized by the maps of specific parameters obtained in Contact, Tapping and tunneling current (TUNA) modes.

3.

Results and Discussion

Formation of big combined biomolecular structures in a buffer solution was verified by photoluminescence experiments during which amyloid specific dye (thioflavin T) was mixed into the colloidal solution. Thioflavin T related increase in the intensity of fluorescence was detected for the solutions with both GDH– Aβ40 and Trx-Aβ40 hybrid proteins compared to the solutions with single biomolecular components. The intensity increase was obtained after incubation of protein solutions at 310 K in PBS buffer (pH = 7.4) for two days. It was supposed that this period was required for formation of fibrils because the increase in the intensity of fluorescence was comparable with the results obtained for the solution with lysozyme fibrils. Self assemblage of extended biomolecular structures on solid surfaces was verified by the SPM measurements. Supramolecular structures were detected on solid surfaces of practically all samples. Aiming to characterize transferability of the structures from the solution onto solid surface, we performed experiments with sufficiently stable lisozyme fibrils. In these experiments hen egg lysozyme fibrils were immobilized on several types of solid substrates listed in section 2 by deposition of the structures from buffer solution. Typical results of these tests are

β-Sheeted Amyloid Fibril Based Structures for Hybrid Nanoobjects on Solid Surfaces

63

illustrated in Fig. 1 by the SPM images obtained for the structures on the surfaces of In2O3 and mica. The same solution was used for deposition of fibrils in the experiment. The surfaces of mica (see Fig. 2b) and SiO2/Si (not illustrated in this report) substrates were found the most favorable for immobilization of the fibrils. In contrast to this, the surfaces of thin In2O3 films seemed more favorable for self arrangement of ball like objects than fibrils as it is illustrated in Fig. 2a.

(a)

(b)

Fig. 1. SPM images of topography of 10 × 10 μm area for self assembled combined structure based on lysozyme fibrils on surface of (a) In2O3 thin film and (b) mica.

It also follows from these results that the density of immobilized fibrils significantly depends on the type of the solid surface. Clearly higher density of fibrils was obtained on mica than Si. Thin SnO and In2O3 films were covered with comparatively small number of fibrils and far between. It can be noted here that frequently it was difficult to detect the rare fibrils on SnO due to significant roughness of the surfaces of these polycrystalline films. Analysis of the SPM force curves on probed substrates revealed that binding forces are slightly greater on mica surfaces that on the rest substrate materials. On the other hand, the binding force on the SiO2/Si surfaces was lower than on SnO films. In spite of this, the density of fibrils was clearly higher on SiO2/Si surfaces than on SnO. Based on the TUNA tests it was supposed that insulating surface is more favorable for immobilization of fibrils than conductive one even if adhesive forces are characterized by the vise versa proportion on these surfaces. Additional experiments are required for more explicit interpretation of dependences of the arrangement of fibrils on the material of substrates. It was proved experimentally that immobilization period can be effectively used for variation of the density of the supramolecular structures on the surfaces. Influence of the immobilization time on the arrangement of the combined structures is illustrated in Fig. 2 by typical SPM images obtained for hen egg lysozyme fibrils deposited on mica surfaces. After deposition of the lysozyme fibrils under similar conditions, the fibrils were rare on the surfaces if immobilization was shorter

64

V. Bukauskas et al.

than 3 minutes while highly dense fibril coating was obtained after period longer than 10 minutes. Moderate and well controlled density of the coverage was obtained between 3 and 10 minutes. Base on the experiments with lysozyme it seems reasonable to suppose that fibrils are transferred from liquid onto solid surface without significant modifications. Dimensions and shape of fibrils seems independent on the origin of the surface and immobilization period. It should be pointed here that some solid surfaces (e.g. In2O3 in our case) and longer immobilization period (>10 minutes) can be assumed being more favorable for agglomeration of fibrils. Assuming formation of fibrils in the solution, self-arrangement of tree-dimensional (3D) strucures on the solid surfaces can be associated with folding, twisting and sticking of fibrils. Immobilized fibrils are stable on the solid surfaces for at least 3–4 weeks. During this period, several analogous SPM images were obtained for each tested sample by separate scans. In general, structures of supramolecular objects of hybrid proteins on the surfaces were individual for each set of fused molecules and different from lysozyme fibrils. Ball-like objects were the most typical objects detected on the surfaces in the experiments with the hybrid proteins. The structure is illustrated in Fig. 3a by typical SPM image. The size of the balls in these structures seemed dependent on the type of hybrid protein but quantitative description of the dependence requires additional studies. Comparatively thick (about 7–14 nm) fibrils were detected in the samples for which the Trx-Aβ40 proteins were used. Length of fibrils was about 1–2 μm. These fibrils were very rare on the solid surfaces. Typical SPM image of such fibril is illustrated in Fig. 3b. Practically continuous granular layer was also visualized in these samples as it is seen in Fig. 3a.

(a)

(b)

Fig. 2. SPM images of 10 × 10 μm area of lysozyme nanofibrils on surface of mica after immobilization period of 3 (a) and 8 (b) minutes. The height (h-scale) is 15 nm.

β-Sheeted Amyloid Fibril Based Structures for Hybrid Nanoobjects on Solid Surfaces

(a)

65

(b)

Fig. 3. SPM images of structures based on GDH-Aβ40 (1.5 × 1.5 μm area, h-scale 6 nm) (a) and a fibril based on Trx-Aβ40 (1 × 1 μm area, h-scale 14 nm) (b) on mica surface.

4.

Conclusions

Hybrid proteins based on Aβ40 peptides can self-arrange in three-dimensional objects that are stable on solid surfaces. Supramolecular structures of hybrid proteins are initially obtained in special solutions by original technology. These structures can be transferred from the solution onto solid surfaces by the method used for immobilization of hen egg lysozyme fibrils. SPM images typically display linear objects of diameter about 2 nm and length about 1–10 μm on dry solid surface after immobilization of the lysozyme fibrils. Agglomeration of the fibrils was detected quite often but this effect was much more frequent for the hybrid proteins. Extended object based on hybrid proteins were larger than lysozyme fibrils. The Trx-Aβ40 hybrid protein formed thick (diameter about 7–10 nm) and short (about 1–2 μm) objects. We think that understanding of formation of the objects can be significantly improved after additional studies.

Acknowledgement. The study was supported by the Lithuanian State Science and Studies Foundation (project ProNanoFiHi) and, in SPI, partly by FP6 project Woundmonitor. References 1 2 3

I. Willner, B. Willner and E. Katz, in Bioelectrochemistry, Vol. 70, 2–11, 2007. T. Liedl, T.L. Sobey and F.C. Simmel, in Nanotoday, Vol. 2, 36–41, 2007. G. Colombo, P. Soto and E. Gazit, in Trends in Biotechnology, Vol. 25, 211–218, 2007.

Characteristics of Vacuum Deposited Sucrose Thin Films F. Ungureanu1, D. Predoi1, R.V. Ghita1, R.A.Vatasescu-Balcan2, and M. Costache 2 1

National Institute of Materials Physics, P.O. Box MG-7, Magurele, Bucharest, Romania, Fax:(040)-21-369 01 77, E-mail: [email protected] 2 University of Bucharest, Faculty of Biology, Molecular Biology Center, Bucharest, Romania Abstract. Thin films of sucrose (C12H22O11) were deposited on thin cut glass substrates by thermal evaporation technique (p ~ 10-5 torr). The surface morphology was putted into evidence by FT-IR and SEM analysis. The experimental results confirm a uniform deposition of an adherent sucrose layer. The biological tests (e.g., cell morphology and cell viability evaluated by measuring mitochondrial dehydrogenise activity with MTT assay) confirm the properties of sucrose thin films as bioactive material. The human fetal osteoblast system grown on thin sucrose film was used for the determination of cell proliferation, cell viability and cell morphology studies.

1.

Introduction

The obtaining of uniform and adherent thin films of sucrose with good biological properties (as maintaining cell morphology in an in vitro system) represents an important goal in the field of biological research. The difference between biomaterials and passive materials consists mainly in their specific biochemical function. Sucrose is a compatible osmolyte, belonging to a class of low molecular weight compounds (C12H22O11) present both in prokaryotic and eukaryotic cells in order to protect proteins against aggressive effects of harsh environmental conditions such as cold and heat stress [1,2]. In this paper we present a study regarding the properties of sucrose thin films deposited on cut glass in medium vacuum conditions, used as culture medium for human fetal osteoblast system. The biological tests confirm the interest regarding the culture medium to be studied in tooth decay, due to the fact that the increasing frequency of sugar application alters dental plaque by reducing its mineral protection capacity [3].

2.

Experimental Methods

2.1. Sample Preparation Powder of sucrose (Merck at 99, 99% purity), was deposited by thermal evaporation using a HOCH VAKUUM Dresden system. The thin films were deposited on cut

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_11, © Springer-Verlag Berlin Heidelberg 2009

67

68

F. Ungureanu et al.

glass substrates. For evaporation in vacuum (p ~ 8 × 10–6 torr) it was used a wolfram boat, and an intensity of the maximum current through boat of Imax ~ 40 A for t ~ 5 sec. The thickness of the sucrose thin films was ~190 nm (sample S1). The “as deposited” sucrose thin films were characterized by different techniques namely: FTIR, SEM together with biological tests. 2.2. Sample Characterizations IR spectroscopic studies were performed in the range 1800–400 cm–1 using a FTIR Spectrum BX apparatus (4000–350 cm–1) in transmission mode with the resolution 8 cm–1. The surface morphology for the deposited sucrose thin films was investigated by scanning electron microscopy (SEM) in a XL-30-ESEM TMP system. Cell morphology and viability were both investigated in this study to assess the biocompatibility of sucrose thin films in an in vitro environment. Cell seeding. Human fetal osteoblasts (hFOB 1.19, CRL-11372, American Type Culture Collection) were seeded at an initial density of 1 × 104 cells cm–2 in a medium containing a 1:1 mixture of Dulbecco’s Modified Eagle’s Medium (DMEM) without phenol red and Ham’s F12 medium and supplemented with 0.3 mg/ml G418, antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin) and 10% foetal bovine serum. Cells were grown for 48 h at 37°C in an humidified atmosphere of 5% CO2 on the cell culture plastic supports (control), and sucrose thin films deposited on glass substrates (S1 substrate). Prior to cell culture, the samples were sterilized by UV light exposure. Cell morphology studies. Phase contrast microscopy was used every 24 h, until the culture reached the confluence, to examine the cell morphology of osteoblasts in contact with the test materials. Substrate dependent changes in cell morphology, density and orientation were evaluated by actin labelling with FITC conjugated phalloidin after 48 h of culture. With this purpose the cells were fixed in 4% paraformaldehyde solution, permeabilised using by 2% BSA/0.1% Triton X-100 and incubated with FITC conjugated phalloidin. Then, the samples were rinsed with PBS and analyzed by microscopy in fluorescence. The cell viability was evaluated by measuring mitochondrial dehydrogenize activity with MTT assay. This assay measures the cell activity, proliferation rate and cell viability. The yellow tetrazolium MTT (3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium bromide) is reduced to insoluble purple formazan granules by all living, metabolically active cells. The precipitated formazan was dissolved in isopropanol, and the absorbance was read at 595 nm. Absorbance values that are lower than the control cells indicate a reduction in the cell activity and viability. Conversely, a higher absorbance rate indicates an increase in cell viability/ proliferation.

3.

Results and Discussion

The deposited thin films of Sucrose were investigated by FTIR spectrometry with the aim to obtain first information about their molecular structure as compared to the powder material (sample S) used for the targets preparation (Fig. 1). The IR spectra for S and S1 samples show the vibration modes of sucrose. In these spectra

Characteristics of vacuum deposited sucrose thin films

69

the characteristic bands of sucrose are observed in the 1250–800 cm–1 range both for S and S1 and from this point of view the exposed spectra are similar [4]. As presented in literature [5] the most suitable region for the IR measurements of sucrose has been found to be the 1250–800 cm–1 region. Namely, the shoulder at 800 cm–1 of sample S can be related to CH2 group and the range 950–1300 cm–1 is related to vibration mode of C-O-C group. The vibration range related to hydrogen bonded water molecules adsorbed on the surface is present in the region 1600 cm –1 for the sample S (as can be observed in Fig. 1). The bands at 1750–1850 cm –1 are related to a C = O bond. The IR spectra of the films (sample S and S1) were recorded in ATR (attenuated total reflection) regime. The difference in intensity of different transmission peaks are related to the thickness of sucrose thin films as the measured exposed volume decreases.

Fig. 1. The FT-IR spectra of sucrose powder (sample S) and thin films (sample S1).

We present in Fig. 2 the SEM micrographs for Sucrose powder (sample S) and Sucrose thin film deposited on glass (sample S1). In order to record a SEM image the Sucrose powder was deposited on a double adhesive carbon band and afterwards was deposited a fine layer of gold (the sample S becomes conductive). We remark on sample S a disordered aspect with scratches and valleys on a uniform background specific for non-crystalline samples. For the Sucrose thin film (sample S1 with a thickness of 190 nm) we remark an ordered aspect of droplets in a uniform matrix.

Fig. 2. The SEM images of sucrose powder (sample S) and thin films (sample S1).

One of the purposes of this study was to investigate the effects of sucrose thin films deposited on glass substrates on cell morphology and viability, in an in vitro human fetal osteoblast system. Comparative morphological examination of the cell monolayer adhered to the plastic support (control) and sucrose thin films (S1), by phase contrast microscopy, showed no differences in cell shape and density.

70

F. Ungureanu et al.

In both cases, the cells displayed a typical elongated phenotype (Fig. 3A, 3B). The localization of actin was examined and correlations between the cytoskeletal organization and morphology of the cell were evaluated. Phalloidin binds to actin filaments much more tightly than to actin monomers, leading to a decrease in the rate constant for the dissociation of actin subunits from filament ends, which essentially stabilizes actin filaments through the prevention of filament deploymerization [6]. Overall, phalloidin is found to react stoichiometrically with actin, strongly promotes actin polymerization, and stabilize actin polymers [7]. In our study, the cells showed an organized actin network dispersed throughout the cell on both analyzed surfaces (Fig. 4A, 4B). From these experiments we observed that the thin films of Sucrose are an appropriate environment for osteoblast cells proliferation. The difference between the osteoblast cells attached to Sucrose in sample S1 medium relative to control osteoblast culture is not obvious. This fact proves that Sucrose thin films are an appropriate medium to be used together with polysaccharides and iron oxides in biological active medium.

Fig. 3. Appearance in phase contrast microscopy of hFOB 1.19 at 24 h of culture: (A) on plastic dishes (control); (B) on the sucrose thin film deposited on glass support (S1). Original magnification 16x.

Fig. 4. Actin cytoskeletal organization (labeled with phalloidin-FITC) of hFOB 1.19 grown for 48h on: a) plastic support (A) – original magnification 16x; b) sucrose film deposited on glass support S1 (B) – original magnification 32x.

Characteristics of vacuum deposited sucrose thin films

4.

71

Conclusions

The sucrose thin films vacuum deposited on glass presented a good adhesion and for in IR analysis it is observed that the band intensity decreases when the thickness of sucrose thin films decreases. The in vitro studies concerning cell morphology and viability displayed by the foetal osteoblast cell line hFOB 1.19 suggests a high biocompatibility of the studied sucrose thin films. They present the feature of preserving the function and structure of osteoblast cells. It is important to conclude that S1 sucrose thin films have micro cells configurations that allow them to be used for obtaining medical biocompatible supports. Acknowledgements. The authors thank to Romanian Scientific Program PNCD II (71-097 and 71-037)/2007 for financial support.

References 1 2 3 4 5 6 7

P. Cioni, E. Bramati, G.B. Strambini, Biophysical Journal, Vol. 88 (2005), 4213–4222 L.C. Provinata, Y. Tou, R.D. Ludescher, Biophysical Journal, Vol. 88 (2005), 3551–3561 E.I.F. Pearce, C.H. Sissons, M. Coleman, X. Wang, S.A. Anderson, L. Wong, Caries Res. Vol. 36 (2002), pp. 87–92 R. Jantas, B. Delczyk, Fibres & Textiles in Eastern Europe, Vol. 13, No. 1(49) (2005), 60–63 F. Cadet, B. Offmann, J. Agric. Food Chem, Vol. 45, No. 1 (1997), pp. 166–171 J.A. Cooper, Effects of Cytochalasin and Phalloidin on Actin. J. Cell Biol. 105 (4): 1473–1478, 1987 J. Wehland, M. Osborn, and K. Weber, Phalloidin-induced actin polymerization in the cytoplasm of cultured cells interferes with cell locomotion and growth. Proc. Natl. Acad. Sci. Vol. 74(12): 5613–5617, 1977

Electropolymerization of Polypyrrole Films in Aqueous Solution with Side-Coupler Agent to Hydrophobic Groups H.M. Alfaro-López1,2, J.R. Aguilar-Hernandez1,*, A. Garcia-Borquez1, M.A. Hernandez-Perez3, and G.S. Contreras-Puente1 1

E.S.F.M. – Instituto Politecnico Nacional Edificio No. 9 U.P.A.L.M. Lindavista C.P. 07738, Mexico D.F., México * E-mail: [email protected] 2 E.S.I.M.E.- I.E – Instituto Politecnico Nacional Edificio No. 2 U.P.A.L.M. Lindavista C.P. 07738, Mexico D.F., México. 3 E.S.I.Q.I.E. – Dpto. Ing. Metalurgica – Instituto Politecnico Nacional U.P.A.L.M. Lindavista C.P. 07738, Mexico D.F., México Abstract. A preliminary study of the electrochemical synthesis and optical characterizacion of the conducting polymer polypyrrole (PPy) was carried out, in order to understand the in-situ electropoliymerization of PPy. Electropolymerization was performed in a three electrode cell by using freshly prepared monomer solutions in presence of a side-coupler agent to hydrophobic groups: sodium dodecylsulfate (DDS), in order to improve the adherence polymer film to the surface of the working electrode. The adherence of the polymer film promotes either the ionic or electronic transport at the interface solution-working electrode. A polar chemical reactive, sodium tetrafluoroborate, TFB, was also used. In order to taylor and optimize the physical properties of the PPy films we varied some parameters during the electropolymerization: the monomer concentration, the electrolyte concentration, pH of the solution and the cell potential. This allowed us to control the oxidation level (impurification or doping) of the polymer. The obtained PPy films were characterized by using UV-Vis and IR spectroscopy. Moreover through scanning electronic microscopy we were able to observe well developed helical structures of polypyrrole.

1.

Introduction

Conducting polymers, like polypyrrole, are semiconducting materials which can be easily processed into thin films for application in organic electronic devices. The physics of these polymers is important, since the disposition of excited state energy governs the efficacy of the polymers as active elements in different devices [1]. Semiconducting polymers have non-degenerated states and in this case the change from double to single bond gives rise to electronic structures with different energy levels. The main difference between semiconducting conjugated polymers and inorganic semiconductors is due to the structural conformation. Conjugated polymers are flexible in nature due to the flexibility of the polymer chains, which also influence positively the exchange of electrical charges [2–3]. Starting from the monomer, pyrrole, the electropolimerization gives rise to the formation of the isomer linked at α positions (α−α). The respective reaction H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_12, © Springer-Verlag Berlin Heidelberg 2009

73

74

H.M. Alfaro-López et al.

mechanism is depicted below showing each one of the reaction steps: a) oxidation, b) dimerization, c) deprotonation and oxidation, d) formation of oligomers and deprotonation, e) oxidation f) propagation or g) overoxidation.

In this kind of materials the excess of electrical charge will occupy localized energy stated in the band-gap. If an electrical charge is added to the polymer chain, it will slowly wander to a localized state. This also will originate a local deformation of the chain, giving rise to an increase of the elastic energy of the system. Usually the additional charge to the polymer chain is not a single electron but a charged radical (cation or anion). This local distortion together with the excess of electrical charge produces localized electronic states in the gap region. This state is usually called polaron, a term borrowed from condensed matter physics [4]. Electrical, optical as well as structural properties of conducting polymer, particularly PPy, depends at great extent upon the growth condition [2–4].

2.

Experimental Methods

Electropolymerization of PPy films was carried out in aqueous solutions in a three electrode cell, together with a Ag/AgCl reference electrode provided with a 1 M KCl junction. All the solutions were freshly prepared and purged with nitrogen during 10 minutes, just before electroplymerization. Pyrrole monomer was reagent grade obtained from Aldrich. As working electrode gold sputtered films onto fused quartz and microscope glass slides were used, which was placed in front of the platinum foil counter electrode. An EG&G Princenton Applied Research (PAR) Potentiostat/Galvanostat model 273 semi-automatically controlled was used for the electropolymerization by applying a constant voltage of 0.6 V against the Ag/ AgCl reference electrode. After deposition PPy films were rinsed with deionized water and dryed with nitrogen. Films thickness was measured with a DEKTAK DK2 profilometer. Measured thicknesses of the films were 10 μm in average. Two different chemicals were used in order to introduce the counter-anion into the polymer chains: sodium dodecylsulfate (DDS), pH = 8, which works also as a surfactant agent and improves the adherence of the PPy film to the substrate. The other salt employed was sodium tetrafluoroborate (TFB), pH = 6, both of them at 0.1 M. Optical characterization in the UV-Vis region was performed with a full automatized Lambda 35 Perkin-Elmer spectrophotometer between 275–1100 nm. IR measurements were recorded with a FTIR-System 2000 Perkin-Elmer spectrometer

Electropolymerization of Polypyrrole Films

75

in specular reflectance mode, with a MCT (mercury-cadmium telluride) detector, between 400–2000 cm–1, i.e. in the so called “finger print” region.

3. Results and Discussion The electronic characteristics of the PPy films was determined through the absorption spectrum. Figure 1 shows the absorption spectra of a couple of PPy films: one of them grown by using the tensoactive agent (DDS): PPy-DDS and the other one grown with TFB: PPy-TFB. In both cases a well defined broad band, with maximum around 3.8 eV, can be observed. On the low energy side a couple of structures around 1.45–1.50 and 1.95 eV can be observed. All of these absorption bands are shifted to higher energies due to the doping level, as compared to a fully doped PPy film [5].

Fig. 1. Visible absorption spectra for PPy films: DDS solid-line and TFB dotted-line.

A fully doped PPy film usually shows two main absorption bands at 1.40 and 2.6 eV, which are related to the transition of valence band to the bonding level of the bipolaron state [6]. The band at 2.6 eV can be considered as the envelope which includes transition from the valence band to the antibonding level of the bipolaron and polaron states [7,8]. Spectroelectrochemical studies of PPy films have demonstrated that the ratio of the absorbance bands at 1.4 and 2.6 eV (γ = I1.4/I2.6 ), mirrors the doping level of the polymer: the higher the ratio γ the lower the doping of the polymer [9,10]. Thus according to this fact, and because band B appears as a shoulder in both spectra of figure 1, the ratio γ should have a high value, then both PPy-DDS and PPy-TFB films are slightly doped due to the

76

H.M. Alfaro-López et al.

overoxidation, which also gives rise to a reduction of the conjugation length. Nevertheless both films are electrochemically active due to the presence of the broad band at 3.86 eV, which is related to the πÆπ* interband transition, i.e. a transition from valence band to conduction band, which means that the band gap of both films is about 3.86 eV. Overoxidation of the studied films was confirmed by infrared measurements. Figure 2 shows the IR spectra of each film, spectra are up shifted in order to be able to compare them. Besides the usual IR bands, overoxidation is confirmed in both samples due to the presence of the carbonyl group ( C = O ) around 1700 cm–1 for the PPy-DDS sample and at 1833 cm–1 [13] for the PPy-TFB sample [11]. The formation of the C = O radical can be explained by considering the fact that pyrrole is a heterocyclic compound, considered as aromatic because of the delocalisation of the π-electrons wich stabilize the ring. These delocalized π-electrons are very reactive and they can promote aromatic electrophilic substitution, wich produces to nitration reactions. Usually aromatic nitriles are obtained by means of nitrogen salts (N2+), wich comes from aromatic amines, pyrrole in our case. The final step of the overall reaction is the production of the radical carbonyl. However, the polymerization of conductive PPy it must be avoided. Plays de role of TFB as electrolyte acts as activator of the reaction. Ar – NH Æ Ar - N2+ Æ omatic amine nitrogen salt

Ar – C ≡ N Æ Ar – COH nitrile carbonyl

(2)

According to the IR measurements and the presence of the carbonyl group, the overoxidation degree must be higher for the PPy-DDS than for the PPy-TFB. All vibrational modes for both films are summarized in Table 1.

Fig. 2. FTIR- spectra of the PPy films: DDS solid-line, and TFB dotted-line.

Electropolymerization of Polypyrrole Films

77

Table 1. IR-vibrational modes of the PPy-DDS and PPy-TFB samples

PPy-DDS Group C=C pyrrole ring C4H5N C-N carbonyl C = O aromatic ring PPy-TFB Band Group a double bonds b aromatic ring c C=N d pyrrole ring C4H5N e C-N f carbonyl C = O Band a b c d e

Wavenumber (cm–1 ) 800/924 1042 1218 1286/1700 1570 Wavenumber (cm–1 ) 670/1000 775 1020 1104 1411 1833

Overoxidation will inhibit to a certain extent further growth of the polymer film due to the lost of electrical conductivity. Conducting PPy polymer films usually show a three dimensional homogeneous couliflower-like growth, as recorded either by scanning electron microscopy or atomic force microscopy [12], as shown in Figure 3, right image for the PPy-TFB sample. However, for the PPy-DDS sample the homogeneous growth is stopped due to the high resistivity of the polymer chains in some parts of the film. In the case the polymer chains tend to roll in a kind of helical growth, figure 3, left image. Moreover on the surface of the substrate cilindrical-like tubes, with diameter of the order of 250 nm, can be distinguished. For the sample PPy-TFB, grown in acid medium, there is no total overoxidation of the polymer chain, the IR band corresponding to the carbonyl group is much less intense, due to the interaction of the carbonyl group with the NH- functional group.

Fig. 3. SEM images of the surface of the a) PPy-DDS films (helical-like growth), left, and b) PPy-TFB, (couliflower-like growth) right.

78

H.M. Alfaro-López et al.

The formation of the C = N radicals is probably due to two mechanisms: on the one hand, to the protonation of the NH- radical, because it has no paired electrons. On the other hand, due to the acid medium and the presence of another kind of free radicals. This facts is corroborated by the presence of the IR band, C = N, around 1411 cm–1 .

4.

Conclusions

Polypyrrole (PPy) films were electroplymerized by using sodium dodecylsulfate (DDS), pH = 8, and sodium tetrafluoroborate (TFB), pH = 6, in aqueous solution. Both kind of polymer films showed absorption bands due to polaron and bipolaron electronic states. The PPy-DDS films has a higher overoxidation grade than the PPy-TFB one, as observed from presence of the carbonyl band at 1700 cm–1 in the IR spectrum. The nucleophilic attack of the carbonyl group by the NH- compound gives rise to the formation of C = N radical, as shown in the respective IR spectrum. SEM images show an homogeneous growth for the PPy-TFB films, whereas helical structures are seen for the PPy-NaDDS, due mainly to the overoxidation of the polymer chains. Acknowledgements. One of us H.M.A.L thanks for the financial support of CONACyT-Mexico for a Doctoral scholarship. J.R.A.H, A.G.B and G.S.C.P COFAA, SNI, EDI fellows. Work partially supported by IPN-SIP-20070990 and CONACyT-52972

References 1 2 3 4 5 6 7 8 9 10 11 12 13

H. Nalva (Ed.), in Organic Conductive Molecules and Polymers, Vol. 1, John Wiley & Sons, Chichester, England 1997. H.S. Nalwa (Ed.), Handbook of Advanced Electronic and Photonic Materials and Devices, Academic Press, San Diego, 2001. T.A. Skotheim, R.L Elsenbaumer, J.R. Reynolds (Eds.), Handbook of Conjugated Polymers, Marcel Dekker, New York, 1996. W.R. Salaneck, I. Lunstrom, B. Ranby, Conjugated Polymers and Related Materials, Oxford Science Publications, Oxford 1993. O. Chauvet, S. Paschen, L. Forro, L. Zuppiroli, Synth. Met., 63, 115, 1994. Y. Li, R. Qian, Synth. Met., 26, 139, 1988. D. Kim, J. Lee, D. Moo, C. Kim, Synth. Met., 69–71, 471, 1995. E. Genies, J. Pernaut, J. Electroanal. Chem., 191, 1515, 1985. T. Lewis, G. Wallace, C. Kim, D. Kim, Synth. Met., 84–86, 403, 1997. K. Yakushi, L.J. Lauchlan, T.C. Clarke, G.B. Street, J. Chem Phys., 79, 4774, 1983. F. Beck, P. Braun, M. Oberst, Ber. Bunsenges. Phys. Chem., 91, 967, 1987. Q. Pei, R. Qian, Synth. Met., 45, 2123, 1991. http://www.cem.msu.edu/reusch/VirtualText/Spectrpy/InfraRed/infrered.htm

Surface Modification of Polymer Powders by a Far Cold Remote Nitrogen Plasma in Fluidized Bed Lynda. Aiche1, 2, Hugues. Vergnes2, Bernard. Despax1, Brigitte. Caussat2 and Hubert. Caquineau1 1

Laboratoire des Plasmas et Conversion des Energies, 118 Route de Narbonne, 31106 Toulouse E-mail: [email protected] 2 Laboratoire de Génie Chimique, UMR CNRS 5503, ENSIACET/INPT, 5 rue Paulin Talabot, BP 1301, 31106 Toulouse Cedex 1, France E-mail: [email protected] Abstract. In this work, nitrogen was grafted on the surface of polyethylene powders in a fluidized bed coupled to a nitrogen microwave post-discharge, under low pressure (10Torr) and low temperature (<90°C). The influences of treatment-duration (1 to 9 h) and nitrogen flow-rate on the XPS (X-ray Photoelectron Spectroscopy) N/C atomic ratio and on the powder wettability have been studied.

1.

Introduction

Nowadays, powders represent about 75% of raw materials used in the industry. One of their main characteristics is their high surface to volume ratio. Therefore, modifying their surface properties is a great scientific and industrial challenge. To treat thermo-sensitive powders, some authors [1–2] have coupled two technologies: a cold plasma to generate active species in a gas and a fluidized bed to promote the gas-particles contact. Nitrogen cold remote plasmas allow grafting nitrogen on surface and then for instance, to increase the hydrophilicity of polymer powders [1–3]. In the present work, the influence of the fluidization ratio and of the run duration has been investigated for the treatment by a cold remote nitrogen plasma of PE powders in a fluidized bed reactor. The fixed bed height to reactor diameter ratio is 1 which corresponds to the definition of a fluidized bed.

2.

Experimental

A scheme of the experimental set-up is presented in Fig. 1. The glass fluidizing column is 50 mm in diameter and 0.9 m in height. Particles were supported by a polypropylene (PP) porous plate. The reactor was maintained under reduced pressure (fixed at 10 Torr) by means of a vacuum pump.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_13, © Springer-Verlag Berlin Heidelberg 2009

79

80

L. Aiche et al.

Before entering the fluidizing column, nitrogen underwent a 2.45 GHz microwave discharge by passing through a microwave cavity (surfatron). The fluidized bed was 25 cm downstream of the surfatron. The power was maintained at 300W. A high density polyethylene (PE) powder, 280 µm in Sauter diameter and 930 kg/m3 in grain density, was fluidized by nitrogen. 51 g of particles were treated per run. The N2 flow rate was varied between 267 and 1000 sccm (standard cm3/min). The treatment duration was varied between 1 and 9 h. Plasma was characterized using Optical Emission Spectroscopy (OES). The OES optical fiber was placed 5 cm above the porous plate. XPS analyses allowed measuring changes in the chemical state and composition of the PE surface induced by the plasma treatments. To pump Fluidized bed reactor D = 50mm H = 90cm

Differentia l Pressure Sensor

Pirani gauge Absolute Pressure Sensor

Fluidized bed Nitrogen Porous plate

Resonant cavity

Micro-waves Plasma

Mass flow regulation

Micro-waves Generator

Fig. 1. Experimental set up.

3.

Results and Discussion

First, the minimum fluidization velocity Umf was deduced from the variation of the pressure drop through the powder bed versus the gas fluidization velocity by applying the the Harrison and Richardson method [4] The experimental Umf at 10 Torr was 3.8 cm/s. The choice to work under a pressure of 10 Torr was a compromise between the concentration of the active species which increases when the pressure decreases and the fluidization quality which was better when the pressure increases. Two kinds of emission are observed using OES. First, the first positive of N2 emission showed up. Its main contribution is at 580 nm and corresponds to the mechanism (1) [5]:

Surface Modification of Polymer Powders

81

N (4s ) + N (4s ) + N ⎛⎜ X 1 Σ + ⎞⎟ → N ( B 3 Π ) + N ⎛⎜ X 1 Σ + ⎞⎟ 2⎝ 2 2⎝ g⎠ g g⎠ N ⎛⎜ B 3 Π ⎞⎟ → N ⎛⎜ A3 Σ + ⎞⎟ + hv(580nm) 2⎝ 2⎝ g⎠ v⎠ N ⎛⎜ A3 Σ + ⎞⎟ + N ⎛⎜ X 1Σ + ⎞⎟ → N ⎛⎜ 4 S ⎞⎟ + N ⎛⎜ 4 S ⎞⎟ + N ⎛⎜ X 1 Σ + ⎞⎟ 2⎝ 2⎝ 2⎝ g⎠ V⎠ g ⎠v ⎝ ⎠ ⎝ ⎠

(1)

As N2(B) results from N(4S) recombination, this emission is also indicative of concentration of N(4S) which is the main reactive species in yellow far cold N2 remote discharges. The other emissive specie is CN with a main contribution at 385 nm corresponding to the radiative transition (2). CN probably results from the N(4S) etching on polymer surface. As the flange maintaining the column is made of polymer, CN emission is even observed without the polypropylene porous plate CN (B, v `' = 7 ) → CN ( X , v '' = 7 ) + hv (385 nm)

(2)

The N2 and CN intensities were followed in the following conditions: 300 W, 1000 sccm N2 and 10 Torr It can be observed in Fig. 2 that: – without the PP porous plate in place, the N2 intensity shows a small decrease between 0–40 min then a plateau is reached at 6500 a.u – With the PP plate, both N2 and CN intensities reach 0 after 40min of treatment. This result is due to an increase of the temperature in the set-up which caused a decrease of the porosity of the PP plate and then an enhancement of the pressure under the distributor. This increase leads (i) to a partial thermalization of the vibrational N2 states which could be unfavorable to N(4S) production (ii) and to a limitation of the propagation of the post-discharge. – In the presence of PE powders, the N2(B) intensity decreases slowly from 1000 to 650 a.u and the CN emission disappear after 3 h of treatment. In this case, the increase of the set-up temperature is much less important than without powders. Therefore, more N(4S) atoms are produced and their transfer through the PP plate is easier. So, the 580 nm emission decrease is much less pronounced. – At the beginning of the experiment, the CN intensity is lower without the powders than with the powders. This can be surprising since CN results from polymer etching and adding powders implies more polymer surfaces in the reactor. Nevertheless, the presence of ramification in the PP structure may ease etching on PP when compared to PE etching. Besides, whereas CN decreases down to zero somehow indicating the end of etching, the decrease of N2 (B) is less important indicating that N active species nevertheless are still present. By analogy with Mafra et al. [6], this might be a consequence of the slow temperature increase which was recorded when the powders are in the reactor. Indeed, they proposed, in their study of the action of atomic oxygen on hexatriacontane, that a temperature increase favors grafting over etching.

82

L. Aiche et al. 6,0E+04

1,6E+03 N2 without PP porous plate

1,4E+03

CN without PP porous plate N2 with PP porous plate

4,0E+04

CN with PP porous plate

3,0E+04 2,0E+04

1,0E+03 8,0E+02 6,0E+02 4,0E+02 2,0E+02

1,0E+04

0,0E+00

0,0E+00 0

(a)

N2 CN

1,2E+03 Intensity(a.u)

Intensity (a.u)

5,0E+04

20 40 Treatment time(min)

0

60

(b)

50

100

150

200

Treatment duration (min)

Fig. 2. Intensities of N2 (580 nm) and CN (385 nm) lines head versus the plasma post discharge duration (a) without powders, with and without porous plate (b) with HDPE powders.

Concerning the powder surface analysis, the as-received PE powder surface presents a single intense C 1s peak centered at 285 eV. After treatment, a N1s peak centered at 401 eV appears. It can be seen in Fig. 3(a) that the N/C ratio measured by XPS increases both with the N2 flow rate and with run duration because of a concomitant increase of the amount of reactive species crossing the bed. By considering an error of measurement of 10%, Fig. 3(b) shows that this ratio almost reaches a plateau after 3 h of treatment, at a value close to 15–17%.

Fig. 3. (a) N/C atomic ratios versus N2 flow rate for 1 and 3 h of treatment. (b) N/C atomic ratios versus run duration for 1000 sccm of N2.

The wettability of the PE powders was evaluated in terms of contact angle using the Washburn method [7] with benzyl alcohol and ethylene glycol. Each contact angle value corresponds to the mean value of three measurements. The accuracy of the results is ±3°.

Surface Modification of Polymer Powders

83

Figure 4 shows that the contact angles decrease with the increase of N2 flow rate and run duration. The best wettability (0°) is obtained with benzyl alcohol for 1000 sccm of N2 and after 3 h. For the same operating conditions, the ethylene glycol contact angle is only of 68°. Just as XPS results, contact angles reach a plateau after 3 h. This levelling corresponds to the competition between functionalization and degradation phenomena; this is a well-known process observed in plasma processing of polymers [3].

Fig. 4. Influence of the N2 flow rate and of treatment duration on the benzyl alcohol (BA) and on the ethylene glycol (EG) contact angles An additional run was performed adding 1 vol. % of O2 to 1000 sccm of N2 for 3 h. It allowed reaching the best wettability of the whole runs.

4.

Conclusions

The N/C ratio and the wettability of powders both increase with run duration and fluidization ratio till reaching a plateau after 3 h. Studies are now in progress to deposit continuous thin layer of SiO2 on the powders surface at low temperature.

References 1 2 3 4 5 6 7

C. Vivien, C. Wartelle, B. Mutel, J. Grimblot, Surf. Interface Anal. 34, 575–579 (2002). B. Leroy, N. Fatah, B. Mutel, J. Grimblot, Plasmas and Polymers 8, 13–29 (2003). F. Bretagnol, M. Tatoulian, F. Arefi-Khonsari, G. Lorang, J. Amouroux, Reactive Functional Pol. 61, 221–232 (2004). J.F. Davidson, D. Harrison, Fluidized particles, Cambridge University Press (1963). S. Villeger, M. Sixou, J. Durand, A. Ricard, J. Phys. D: Appl. Phys, 39, 3826–3830 (2006). M. Mafra,T. Belmonte, A. Maliska, ITFPC 07 conference, Nancy (2007). E.W. Washburn, Phys. Rev. Ser. 17 (3), 273–283 (1921).

Features of Polytetrafluoroethylene Coating Growth on Activated Surfaces from Gas Phase Alexander A. Rogachev1, Sigitas Tamulevičius2, Alexander V. Rogachev3, Igoris Prosycevas2 and Mindaugas Andrulevičius 2 1

Belarusian State University of Transport, Kirov Str. 34, Gomel, 246653, Belarus Institute of Physical Electronics, Kaunas University of Technology, Savanoriu 271, LT-3009 Kaunas, Lithuania 3 Francisk Scorina Gomel State University, Sovetskaya Str. 104, Gomel, 246019, Belarus 2

Abstract. The kinetics and morphological peculiarities of polytetrafluoroethylene (PTFE) coating growth from the active gas phase on differently treated substrates by solvent, N+ or Ar+ ions are considered. Wettability of thin films was tested using water and glycerin liquids for contact angle measurements, furthermore surface energy was also calculated. Chemical composition of the deposited thin polymeric films was analyzed by X-ray photoelectron spectroscopy (XPS), film morphology was studied by atomic force microscopy (AFM). An additional mathematical analysis of the cluster formation images was carried out. It was found that the substrate surface energy effects on the distribution of polymer microparticles both at the initial stage as well as during coating growth. The substrate surface activation by ion beam results formation of a continuous low-molecular coating with a thinner apparent thickness.

1.

Introduction

Thin polymer coatings obtained by vacuum methods have many applications in electronics, optics, mechanical engineering and medicine. They have an outstanding combination of chemical and physical properties such as superior chemical stability, high thermal stability, low friction coefficient and low surface tension. Moreover, PTFE coatings have excellent electrical properties, low dielectric constant and dielectric loss tangent, high insulation and a breakdown voltage. Protective and hydrophobic properties of PTFE coatings depend first of all on the uniformity of film, especially in the case of very thin layers [1–3]. The present work is aimed at studying the kinetics, structural and morphological regularities of the onset and initial growth of PTFE coatings on crystalline silicon substrate surfaces exposed to various pretreatment procedures.

2.

Experimental Methods

Variations in morphology and adsorption properties of PTFE coatings along with the effect the substrate pretreatment have been studied at the initial stages of coating formation. PTFE coatings have been deposited from the active gas phase obtained by the electron-beam dispersion of the original powder of polymer in vacuum [4].

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_14, © Springer-Verlag Berlin Heidelberg 2009

85

86

A.A. Rogachev et al.

The coating deposition of a required thickness gradient was reached by using a computer-controlled shutter. The growth rate was 0.5–1 nm/sec and it was measured with a quartz crystal microbalance (QCM) [5]. The substrates were single-crystalline silicon wafers treated with organic solvent (a mixture of aromatic hydrocarbons, ketones, alcohol and ethers) and a beam of nitrogen (N+) or argon (Ar+) ions. The ion energy constituted 3 keV and ion current density J = 1.2 A/m2. The exposure time was less than 5 min. To study the surface properties of the coatings, the wetting angle with distilled water and glycerol was measured. The polar and dispersion components of the surface energy were determined by the Owens Went method [6]. The XPS measurements were done on the XSAM800 spectrometer (Kratos Analytical Ltd, UK) with a Mg Kα x-ray source (1253.6 eV photons). Surface morphology of the treated films was studied by the AFM. An additional mathematical analysis of the cluster formation images was carried out. The cuttingplane method and cluster labeling [7] were employed to plot the distribution of islets over the area n(S). Characteristic sections with the maximum number of isolated clusters not touching the image boundary were chosen for the analysis.

3.

Results and Discussion

Ion treatment of the single-crystalline silicon wafers changes considerably its surface energy. Most noticeable increase (almost 1.5 times) in the surface energy and especially in its polar component occurs when the single-crystalline silicon is treated with nitrogen ions (Table 1). Table 1. Surface energy of silicon

Surface energy components, mJ/m2 Dispersion Polar Total energy

pristine 3.5 50.6 54.1

Treatment agent solvent solvent-N+ 7.9 9.3 45.2 64.9 53.1 74.2

solvent-Ar+ 7.5 66.2 73.7

The evident differences were recorded in morphology of the produced layers and their properties [8] when the volatile products of PTFE dispersion were deposited on the substrate surface pretreated with the solvent or solvent and ion beam. In the case of PTFE coating produced on the silicon surface treated with a solvent, the adsorption and formation of the polymer phase microparticles was characterized by an essential surface inhomogeneity (Fig. 1). Rather large in size and up to 20 nm high particles were observed during the initial stage of the coating growth with the apparent thickness 2.4…12 nm (Fig. 1a). When the layer thickness reached 7 nm, the character of the surface filling degree over the basal area of the particles changed (Fig. 1b). As a result, the multimodal distribution, occurring for the apparent thickness 2.4 nm, transforms into the unimodal one. This is a proof of conversion of the deposition character at a certain stage of the coating growth. The non-stationary deposition phase, observed during the very onset of the coating formation, is attributed

Features of Polytetrafluoroethylene Coating Growth

87

15 10

20 10

n(S )

n (S )

0.6

0.6 0.4

0.4 0.2 0

0.2

0.1

0.2

0.3 S, μm2

0

a)

0.1

20 10

n(S )

n (S )

0.6

0.6

0.4

0.5

0.2 0

0.1

0.2

0.3 S, μm2

b)

0.2

0.3 S, μm2

d)

6 4

0.2

0.3 S, μm2

c)

0

0.1

Fig. 1. AFM images of PTFE coatings and corresponding distribution patterns of surface filling over microparticle area: a) apparent thickness – 2.4 nm; b) 7.3 nm; c) 12.1 nm; d) 18.3 nm. The silicon substrate surface was treated with the solvent. Scanning area 12.6 × 12.6 μm.

to the continuous generation of the particles, which later transforms into a stage of the dominating lateral growth of the earlier formed particles. Changes in the monocrystal substrate adsorption properties impose an effect on the formation process of the polymer coatings. Ion treatment of the substrate increases the total surface energy largely at the expense of the increasing polar component. In this way, some discrete structures have not been detected at the initial stages of polymer deposition. As the apparent layer thickness reaches 3 nm, the coating becomes continuous. Surface energy of the layer decreases to 14 mJ/m2 in consequence of variations the polar component, typical for the fluoropolymers with high content of nonpolar CF3 groups [9].

88

A.A. Rogachev et al.

1 -CF3 2 CF3-CF23 -(CF2-CF2)4 -CF-CF25 -C-F 6 -C-CF2 7 -C-CF8 -C-C-, -C-H9 conjugated carbons

Intensity, Arb. Units

3 2

3

8

7 8 1

4 5

291 286 Binding Energy, eV

4

281

296

а)

7 9

5 6

1

9

6 296

2

291 286 Binding Energy, eV

281

b)

Fig. 2. The XPS C1s core-level spectra of the PTFE films with thickness 7.5 nm on the silicon surface treated with solvent and ions N+ (a), only solvent (b). The extreme low-energy contribution of the satellite peak was not taken in account.

The chemical structure of the material surfaces was analyzed by XPS. Highresolution scans of the C1s region for the analyzed PTFE films are shown in Fig. 2. The line-shape analysis by peak deconvolution shows that the C1s spectrum for PTFE films contains nine distinct peaks [10]. The relative chemical composition of the C1s spectra for our samples is shown in Table 2. Table 2. Relative chemical compositions of the PTFE films

-C-CF-

-C-C-, -C-H-

Conjugated carbons

9

-C-CF2

8

-C-F

7

-CF-CF2-

Cls components (%) 4 5 6

-(CF2-CF2)-

Solvent-N+ Solvent

3

CF3-CF2-

Pretreatment agent

2

-CF3

1

6.8

24.6

27.5

5.3

6.0

0.9

10.1

12.9

6.0

7.2

7.5

39.2

6.7

6.5

2.8

9.6

13.1

7.3

Comparing the chemical composition of the thin polymeric coating on the silicon substrate pretreated with N+ ions and with the solvent only one can see that there is a relative increase in the content of СF3-CF2- , while the content of -СF2CF2- and С-СF2 groups decreases. The content of -C-C-, CF3 groups and other groups differs less than 0.5%. This experimental fact indicates that the thin PTFE films formed on the silicon substrate pretreated with N+ ions include the linear low-molecular fragments, while the PTFE films on the silicon substrate pretreated with the solvent only include the branching and cross-linking structure. Therewith, high concentration of -СF2-CF2-, С-СF2 groups indicates the relative high length of the molecular branch.

Features of Polytetrafluoroethylene Coating Growth

4.

89

Conclusion

Thus, we have studied the morphological features of the growth of PTFE coatings on the silicon surface activated by different methods (solvent or ion beam) and the adsorption properties of these coating. Ion treatment of single-crystalline silicon wafers by ion N+ or Ar+ changes considerably value of the surface energy and its polar components (by a factor of 1.5). The deposition on these surfaces is characterized by low activation energy and yields continuous layer at a low apparent thickness. These films have a mainly low-molecular, linear structure. On the solvent-treated silicon surface, the polymeric coating is deposited selectively. Adsorption and nucleation of polymeric microparticles on the silicon surface are nonuniform. These polymer films contain cross-links with the relative high length of the molecular branches. Variations in the adsorption properties of the substrate lead to the changes of the mechanism of nucleation, morphology, molecular structure and wetting ability of PTFE films respectively.

References 1. 2. 3. 4.

H. Biederman, Y. Osada, Plasma Polymerization Processes, Elsevier, Amsterdam, 1992. H. Yasuda Plasma polymerization, Academic Press: New York, 1985. Konstantin P. Gritsenko and Anatoly M. Krasovsky Chem. Rev. Vol. 103, 3607, 2003. Structure and properties of nanocomposite polymer coatings A. Rogachev, M Yarmolenko, A. V. Rahachou, S. Tamulevicius, I. Prosycevas. Journal of Physics: Conference Series 100, 082042, 2008. 5. Lu C., Czanderna A. W. Application of Piezoelectric Quartz Crystal Microbalances, Elsevier: Amsterdam, 1984. 6. Lee Lieng-Huang. J. Adhesion Sci. Technol. Vol. 7, 6, 583, 1993. 7. J. Hoshen, R. Kopelman Phys. Rev. B14, 3488, 1976. 8. А. A. Rogachev, Russian Journal of Applied Chemistry. Vol. 77, 2, 281, 2004. 9. L. A. Woll, Fluoropolymers, Wiley-Interscience, New York, 1972. 10. G. Beamson, and D. Briggs, in High Resolution XPS of organic polymers (The Scienta ESCA 300 Database), Wiley, Chichester 1992.

Modification of Amorphous Carbon Film Surfaces by Thermal Grafting of Alkene Molecules Hussein Sabbah1, Abdelkader Zebda1, Soraya Ababou-Girard1, Bruno Fabre2, Stéphanie Députier 3, André Perrin3, Maryline Guilloux-Viry 3 and Francine Solal 1, Christian Godet1 1

Physique des Surfaces et Interfaces, Institut de Physique de Rennes (CNRS UMR 6251), Université Rennes 1, Beaulieu - Bât. 11C - 35042 Rennes (France) E-mail: [email protected] 2 Matière Condensée et Systèmes Electroactifs, Sciences Chimiques de Rennes (CNRS UMR 6226), Université Rennes 1, Beaulieu - Bât. 10C - 35042 Rennes (France) 3 Chimie du Solide et Matériaux, Sciences chimiques de Rennes (CNRS UMR 6226), Université Rennes 1, Campus de Beaulieu, 35042 Rennes (France) Abstract. Thermally-assisted grafting of linear alkene molecules either in the liquid phase (ethyl undecylenate) or in the gas phase (perfluorodecene), has been performed on atomically flat amorphous carbon (a-C) films with variable average surface hybridization, sp3/(sp2 + sp3), as obtained from X-ray photoelectron spectroscopy. In contrast with the sp2-rich sputtered a-C, optimized sp3-rich a-C films obtained by Pulsed Laser ablation of a glassy carbon target do not require surface preparation before liquid phase grafting.

1.

Introduction

Interface control at the molecular level is important for the development of molecular electronic junctions [1,2] with potential impact in the fields of molecular memories, photovoltaic conversion and biochemical sensing. In order to achieve homogeneous and reliable long term electrical properties, the design of robust interfaces requires the selection of molecules bearing a chemical functionality reactive towards a solid surface, usually a metal (e.g., gold) or a semiconductor (e.g., hydrogen-passivated crystalline silicon). However, these materials may not be ideal substrates because thiol chemistry produces weak Au-S bonds (167 kJ/mol), while the unreacted Si-H interface bonds at Si(111):H surfaces are prone to partial oxidation [3]. Alternative strategies address the immobilization of organic molecular layers (OML) on carbon-based surfaces through C-C covalent bonds (348 kJ/mol), using either photochemical reaction of alkenes with H-passivated polycrystalline diamond [4,5] or electrochemical grafting on glassy carbon electrodes [6,7]. Most grafting studies with amorphous carbon (a-C) surfaces were performed with sp2 -rich films [8,9]; the surface preparation required for a significant OML coverage on a-C [8] is clearly a process drawback. In this work, a-C surfaces with variable average surface hybridization of C atoms, sp3/(sp2 + sp3), are compared using a thermally-assisted process used for covalent grafting of linear alkene molecules in liquid phase or gas phase.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_15, © Springer-Verlag Berlin Heidelberg 2009

91

92

H. Sabbah et al.

2.

Experimental Methods

Pulsed Laser Deposition (PLD) a-C films were grown at room-temperature using a KrF laser (Tuilaser Excistar λ = 248 nm, 2 Hz, 40 ns pulse width, fluence of 7.5 J.cm–2). After ultrasonic bath cleaning in acetone, isopropanol and water, crystalline silicon (c-Si) substrates were placed at 60 mm from a rotating glassy carbon target (Sigradur G, HTW). To obtain a-C surfaces free of micron-sized particles ejected from the target, before deposition, the target was cleaned for 5 minutes at 2 Hz with a shutter in front of the substrate. A second set of a-C films was grown at roomtemperature using Magnetron Sputtering of a graphite target in argon/hydrogen mixtures [8]. Liquid phase grafting of ethyl undecylenate CH2 = CH(CH2)8-COOC2H5 [8] and gas phase grafting of perfluorodecene CH2 = CH(CF2)7-CF3 were respectively performed at 160°C and 230°C. XPS data (X-ray source Mg Kα, hν = 1254 eV) provide the OML surface coverage values. Monochromatized excitation (Al Kα, hν = 1486 eV, total resolution of 0.6 eV) is used for the hybridization studies. Surface energies were derived from contact angle measurements using water, diiodomethane and glycerol, after surface rinsing in VLSI grade acetone [10].

3.

Results and Discussion

XPS data [10] show that sputtered a-C is sp2-rich while PLD a-C is sp3-rich (Table 1) in line with Raman and optical data. Both surfaces are suitable for OML grafting (AFM roughness <1 nm). As-received sputtered a-C surfaces being significantly oxidized ([O] = 7–9 at.%) some films were prepared using Ar+ bombardment and thermal annealing. PLD a-C was used as-deposited, after a few minutes in air leading to low surface oxidation [O] = 4–6 at.%. Surface coverage values deduced from O = C-O and CF2 intensities in C1s spectra are summarized in Table 1. For ester grafting on sputtered a-C, since a strong substrate oxidation contributes to the O = C-O signal, the coverage was deduced from the attenuation of the main C1s component (C-C) [8]. A good correlation between increasing PFD coverage and decreasing dispersive surface energy γ LW is observed, while the ester-OML slightly increases γ LW. Table 1. Surface C atom hybridization and OML coverage (±10%) for liquid phase and gas phase grafting on PLD and sputtered a-C films (* after Ar+ surface preparation)

Amorphous Carbon

sp3/ (sp2+sp3) (±0.05)

Ethyl undecylenate (ester) Liquid phase (16 h)

Sputtering

0.18

*[O] = 1 at.%

Pulsed Laser

0.55

Deposition

Perfluorodecene (CFx) Gas phase (2 h)

4.1 × 1014 cm–2

[O] = 8 at.% 2.2 × 1014 cm–2

[O] = 4.5 at.%

[O] = 4.5 at.%

14

–2

3.9 × 10 cm

1.4×1014 cm–2

Modification of Amorphous Carbon Film

Surface energy γ

LW

-2

(mJ m )

F E

50

93

Ester liquid phase on PLD a-C

Sputtered a-C

x

40 30

PLD a-C

Perfluorodecene gas phase on PLD a-C

20 Teflon foil

10 0 0

1

2

3 14

4 -2

Grafted OML coverage (10 cm ) Fig. 1. Dependence of the dispersive surface energy γ LW on the surface coverage of different molecular layers covalently grafted to amorphous carbon surfaces.

4.

Conclusion

In contrast with sp2-rich sputtered a-C investigated previously [8], optimized sp3-rich PLD a-C films do not require surface preparation before liquid phase grafting. Both the high atom density and the sp3-rich surface hybridization are expected to decrease the reactivity of PLD carbon films towards oxidation at the ambient (before grafting) and towards oxidizing impurities in solution.

References 1 2 3

J. Jortner and M. A. Ratner, Molecular Electronics, Blackwell, Oxford, 1997. D. Cahen and G. Hodes, Adv. Mater. 4, 789, 2002. Z. Lin, T. Strother, W. Cai, X. Cao, L. M. Smith, and R. J. Hamers, Langmuir 18, 788, 2002. 4 T. L. Lasseter, B. H. Clare, N. L. Abbott, and R. J. Hamers, J. Am. Chem. Soc. 126, 10220, 2004. 5 W. Yang, O. Auciello, J. E. Butler, W. Cai, J. A. Carlisle, J. E. Gerbi, D. M. Gruen, T. Knickerbocker, T. L. Lasseter, J. N. Russell, L. M. Smith, and R. J. Hamers, Nature Mater. 1, 253, 2002. 6 Y. C. Liu and R. L. McCreery, J. Am. Chem. Soc. 117, 11254, 1995. 7 P. Allongue, M. Delamar, B. Desbat, O. Fagebaume, R. Hitmi, J. Pinson, and J. M. Saveant, J. Am. Chem. Soc. 119, 201, 1997. 8 S. Ababou-Girard, H. Sabbah, B. Fabre, K. Zellama, F. Solal, and C. Godet, J. Phys. Chem. C 111, 3099, 2007. 9 B. Sun, P. E. Colavita, H. Kim, M. Lockett, M. S. Marcus, L. M. Smith, and R. J. Hamers, Langmuir 22, 9598, 2006. 10 A. Zebda, H. Sabbah, S. Ababou-Girard, F. Solal, and C. Godet, Appl. Surf. Sci. 254, 4980, 2008.

DNA-controlled Assemblage of Ag Nanoparticles on Solid Surfaces V. Bukauskas, A. Šetkus, I. Šimkienė, J. Sabataitytė, A. Kindurys, A. Rėza and J. Babonas Semiconductor Physics Institute, Gostauto 11, Vilnius, Lithuania E-mail: [email protected] Abstract. Arrangement of DNA based structures on mica and modified Si surfaces is investigated by methods of scanning probe microscope (SPM) and spectroscopic ellipsometry (SE). DNA strands are deposited from a colloidal solution on the surfaces at room temperature. The surfaces were additionally modified with Ag nanoparticles by special technology. The surface structures are visualized by SPM. The effect of the multicomponent structures on the optical response of complex hybrid structures is studied. The optical response of the hybrid samples is related to the contributions of DNA and Ag nanoparticles on the Si surfaces. Formation of combined structures based on Ag nanoparticles and DNA strands is discussed.

1.

Introduction

During the last decade an interest in the development of self-arranged hybrid structures is increasing rapidly. Advantages of these structures are expected being originated from unique properties of biomolecules such as proteins, desoxyribonucleic acid (DNA) and redox enzymes. Integration of the biomolecules within solid nanostructures introduces biochemical recognition in technology and functioning of new type devices. Components of electronic circuits, sensors and actuators can be created by nanotechnology and self-assembling of supramolecules [1,2]. DNA molecules are the most frequently used “building blocks” for assemblies of various dimensions between nanometers and micrometers [1,2]. The negative charge in phosphate backbone of DNA molecules is favourable for fastening metal nanoparticles and, consequently, for creating multifunctional and multicomponent nanostructures. Self-assembling of biomolecules depends critically on their interaction with the solid surface. The biomolecules can form the covalent bonds by linking to the surface through silanole (-O-Si-), amidine (-NH-(C-O)-), phosphonate (O-(HPO2)-), carboxyle (-O-(C-O)-) [3] and amine (-NH2) [4] groups. Therefore, for enhancing the linkage, the surfaces of glass, quartz or mica substrates are usually modified by these functional groups [5]. Surface of c-Si substrates can also be modified by advanced technologies developed for modern sophisticated electronic devices. It makes c-Si attractive material for creating hybrid surface structures that can easily be integrated within multifunctional clusters of electronic devices. However, development of hybrid structures with biomolecules is often problematic due to complicated c-Si surface

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4–16, © Springer-Verlag Berlin Heidelberg 2009

95

96

V. Bukauskas et al.

chemistry and comparatively high roughness of the surface that frequently limits morphological studies at individual stages of the technology. The present study aims to investigate development of ordered metallic structures by self-assembling of DNA molecules and Ag nanoparticles on modified Si surfaces. Changes in the surface morphology produced by the technology at specific stages were analyzed in special samples on mica substrates with atomically flat surfaces by the SPM modes. The complete structures on the Si surface were studied by combined SPM and optical methods.

2.

Samples and Experimental Methods

Standard Si wafers (n-Si (100), Ø 5 cm, resistance 0.5 Ω cm, surface roughness ~3 nm) were used for preparation of hybrid structures. Cleaned Si wafers were immersed for 4 min into 1% APTES (3-aminopropyl-triethoxysilane) solution in toluene at 60oC. According to ellipsometric data, the thickness of APTES layer was ~10 nm. The hybrid samples were prepared by deposition of DNA and Ag nanoparticles from solution on Si substrates with modified surface. Ag nanoparticles (Ø 5 nm) were covered by polyvynilpropylene (PVP). For the investigation of nanometric objects such as DNA and Ag nanoparticles, mica substrates with atomically flat surface were prepared. Structure of the sample surfaces and electrical properties were investigated by SPM D3100/Nanoscope IVa (Veeco, Digital Instruments). The surface was visualized in tapping mode, distribution of electrical properties were investigated by electric force microscopy (EFM). SE technique was used for characterization of hybrid samples and identification of the components in complex structures due to their contribution to optical response. Photometric ellipsometers with rotating analyzer and photoelastic modulator of light polarization were used in the spectral range 250–800 nm.

3.

Results and Discussion

Typical DNA structures self-assembled on mica surface are illustrated by SPM images in Fig. 1. In Fig. 1a topography of samples displays ball and thread like structures. Thread like structures are typical for the samples with DNA on mica [6]. Height of the DNA is equal to about 7–8 nm and length varies between 0.5 and 2 µm. The main SPM phase (Fig. 1b) reveals differences in adhesion forces and viscoelastic properties of the surface areas. Bright objects in Fig. 1b we attribute to precipitates of salts. The light grey objects that clearly correspond to the DNA in Fig. 1b are supposed being much softer than the bright ones. The origin of these light grey spots that correspond to the ball like objects in Fig. 1a are not determined in present report. More detailed analysis of the SPM phase contrast is not the subject of our present report and will be presented elsewhere. Metallized DNA layer on mica substrate is shown by images of the SPM topography in Fig. 2a. The surface of the sample is covered by comparatively smooth coating (dark background in Fig. 2a) and separated aggregates on the top

DNA-controlled Assemblage of Ag Nanoparticles on Solid Surfaces

(a)

97

(b)

Fig. 1. The SPM images of topography (a) and phase (b) for DNA layer self-assembled from colloidal solution on the mica surface then rinsed in distilled water and dried.

(a)

(b)

Fig. 2. The SPM images of topography (a) and electric force phase distribution (b) for DNA and Ag nanograins on a mica substrate with the base voltage +1 V.

of it (bright spots). Height of the aggregates varies between 7 and 15 nm. The EFM imaging proved the presence of charged domains on the surfaces. The surface potential distribution over the scanned area is illustrated by the EFM phase images in Fig. 2b (base voltage +1 V). The bright spots represent the charged domains and can be associated with clean Ag nanoparticles. Coated and smaller particles are visible if higher base voltage is applied to the substrate. Typical map of electric force phase for the base voltage +10 V is shown in Fig. 3. Metallic nanoparticles that are partly or even completely covered by an isolating coating are explicitly displayed by the bright and grey spots in the EFM phase image if the applied dc-voltage is high enough. Most of these spots are invisible in

98

V. Bukauskas et al.

the height image in Fig. 2a. Thread like structures can be traced in Fig. 3 that can be associated with DNA strands on mica substrate. Changes in the surface structure at different stages of the combined layer technology can be analyzed on very smooth mica substrates. The lines similar to those in Fig. 3a and composed of the grains were also seen in the SPM images of hybrid samples Ag/DNA/APTES/Si (Fig. 3b). The optical response of hybrid samples was studied taking into the presence of DNA and Ag nanoparticles on APTES-modified Si substrates.

(a)

(b) 2

Fig. 3. The SPM images (area 1.1 × 1.1 μm ) of the DNA and Ag nanograins on mica (as in Fig. 2, EMF at the base +10 V) (a) and on APTES-modified Si substrate (tapping mode) (b).

The spectral dependence of the optical response of APTES layer deposited on Si substrate was well described by standard dielectric function of SiO2. However, the weight varied in the range 0.5–1.0 indicating the porosity of APTES layer which is known to be dependent on technological regime of formation. Figure 4a illustrates the spectral dependence of ellipsometric parameters Ψ and Δ for the hybrid sample Ag/APTES/Si. Experimental spectra were fitted by the optical response of one effective layer. According to model calculations for this sample, the thickness of the effective layer and APTES film was 5.3 and 11.5 nm, respectively. The spectral dependence of the dielectric function for the effective layer (Fig. 4b) possesses two features. The low-energy peak at ~2.2 eV can be attributed to the residual material of the solution containing the PVP-coated Ag nanoparticles. The peak can be also contributed by the interparticle dipoledipole couplings of nanoparticles on solid substrates. The peak at the ~3.4 eV is related to the surface plasmon resonance of metal nanoparticles and corresponds to the absorption peak of Ag colloidal solution (Fig. 4b). In the spectra of hybrid samples Ag/DNA/APTES/Si, the peak at ~4.5 eV originated from the contribution of DNA was additionally observed.

DNA-controlled Assemblage of Ag Nanoparticles on Solid Surfaces

(a)

99

(b)

Fig. 4. (a) Experimental (points) and modelled (curves) of ellipsometric parameters Ψ and Δ for the sample Ag/APTES/Si and (b) dielectric function spectra of the effective surface layer obtained from the fitting procedure as compared with the absorption spectrum of colloidal solution of Ag nanoparticles.

4.

Conclusions

The self-arranged ordered DNA structures decorated by Ag nanoparticles are effectively formed on the Si surface modified by APTES. The investigations have shown the commonalities and some specific features in the hybrid structures on smooth mica surface and modified Si substrates. Small number of regular structures on mica surfaces proved that binding force of DNA to the surfaces is much lower in mica-DNA system as compared to the hybrid structure based on modified Si substrates. Spectroscopic ellipsometry measurements have shown the contributions to optical response due to components in hybrid samples. Acknowledgement. The study is supported by Lithuanian State Science and Studies Foundation.

References 1 2 3 4 5 6

I. Willner, B. Willner and E. Katz, in Bioelectrochemistry, Vol. 70, 2–11, 2007. T. Liedl, T.L. Sobey and F.C. Simmel, in Nanotoday, Vol. 2, 36–41, 2007. K. Kalyanasundaram and M. Grätzel, in Coord. Chem. Rev., Vol. 177, 347–414, 1998. S.H. Behrens and D.G. Grier, in J. Chem. Phys., Vol. 115, 6716–6721, 2001. J. Kobayashi, T. Hinoue and H. Watarai, in Bull. Chem. Soc. Jpn., Vol. 71, 1847–1855, 1998. F. Moreno-Herrero, P. Herrero, F. Moreno, J. Colchero and C. Gomez-Navarro, J. Gomez-Herrero and A.M. Bar, in Nanotechnology, Vol. 14, 128–133, 2003.

Characterization of Self Assembled Monolayer Formation of 11-Mercaptoundecanoic Acid on Gold Surfaces Johanna Stettner1, Paul Frank1, Thomas Griesser2, Gregor Trimmel2, Robert Schennach1, Roland Resel1 and Adolf Winkler1 1 2

Institute of Solid State Physics, Graz University of Technology, Austria Institute for Chemistry and Technology of Materials, Graz University of Technology, Austria

Abstract. Self assembled monolayers (SAMs) of 11-mercaptoundecanoic acid (11-MUA) were investigated by means of x-ray photoelectron spectroscopy (XPS), atomic force microscopy (AFM) and thermal desorption spectroscopy (TDS). SAMs were prepared in solution, using Au(111)/mica and recrystallized gold foils as substrates. Different binding energies of the S2p peak measured by XPS as well as different thermal desorption behavior indicate the influence of the substrate on the SAM formation. The stability of the SAM under ambient conditions was investigated in order to determine long term effects, showing rearrangements as well as depletion of the 11-MUA layer.

1.

Introduction

In the last decades, a vast interest in self assembled monolayers of alkanethiols on gold has developed, due to the wide range of possible applications in chemical sensing, corrosion protection and lithography [1,2,3]. There is a consent that methyl-terminated, long alkanethiols (n > 6) form well ordered monolayers on Au(111), with a chain tilt of ~30° with respect to the surface normal. The alkanethiols usually arrange on Au(111) in a ( 3 × 3 ) R30° superstructure [4]. One quite often investigated system is mercaptoundecanoic acid (11-MUA) on Au(111) surfaces, because it affords the opportunity to replace the acid end group by other functionalized groups quite easily [5]. In general, there is good agreement that this system yields a well ordered monolayer under well defined preparation conditions, even though acid-terminated SAMs might exhibit less ordering than simple alkanethiols [6–9]. However, the existence of ordered domains for 11-MUA on Au(111) was demonstrated by means of STM [4,10,11]. But still, there is a lack of information about the thermal stability as well as to the stability of the SAMs under ambient conditions. In this contribution we attempt to provide an insight into the complex processes which may occur on the SAM under prolonged ambient conditions and during thermal desorption of the molecules. An appropriate method for these investigations is thermal desorption spectroscopy (TDS). This method not only allows to determine the desorption temperature of the intact molecules, but also to identify decomposition products on the surface by measuring fragments of the molecule mass with the help of a mass spectrometer.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_17, © Springer-Verlag Berlin Heidelberg 2009

101

102

J. Stettner et al.

2.

Experimental Methods

SAM Formation. SAMs of 11-MUA were prepared ex situ in 1–3 mM ethanolic solution of 11-mercaptoundecanoic acid, which was purchased at Sigma Aldrich. Two different substrates were used: Au(111)/mica and recrystallized gold foils. Typically, substrates are cleaned in piranha solution. This procedure was not possible for Au(111)/mica, because it led to a separation of the gold layer from the mica substrate. Therefore, both substrates were cleaned in a UHV chamber by Ar+-sputtering, followed by short annealing at 800 K to get a smooth surface. After removal from the vacuum chamber, it was immediately immersed into the solution. Typically samples remained for 48 h in the solution. After removal out of solution, the samples were rinsed with ethanol and dried with CO2 – spray. SAM Characterization. TDS and XPS were performed in a UHV chamber with a base pressure of 10–10 mbar. The apparatus is equipped with a x-ray photoelectron spectrometer (Leybold Heraues, EA 10/100), a quadrupole mass spectrometer (Balzers QMA 400) with a mass range from 1 to 500, and an Ar+-sputter gun. For TDS the sample was mounted on a steel plate, which can be heated by tantalum wires spot welded on the backside of the plate [12]. AFM was performed ex situ using a Nanosurf Easyscan 2 scanning probe microscope in tapping mode.

3.

Results and Discussion

X-ray Photoelectron Spectroscopy. An overall XPS spectrum of the 11-MUA film on Au(111)/mica is shown in Figure. 1a, which displays the O1s and the C1s peaks, in addition to the gold peaks. Figure. 1b shows the high resolution spectrum of the C1s peak at 284,7 eV with a broad shoulder up to 289 eV. As reported in several papers, this shoulder can be referred to carboxylic carbon of the acid end group [4, 11]. The S2p signal is so small that it can’t be seen in the overview spectrum (Fig. 1a). Figure. 2 shows long time integrated spectra of the S2p peaks for the SAM on Au(111)/mica and gold foil substrates, respectively.

Fig. 1. Overview XPS – spectrum of 11-MUA on Au(111)/mica (a). The high resolution spectrum of the C1s signal. One can clearly see a broad shoulder at 289 eV, indicating the carboxylic group of the acidic end group of the SAM (b).

Characterization of Self Assembled Monolayer Formation

103

Fig. 2. The S2p peak of 11-MUA on Au(111)/mica (a) and on recrystallized gold(foil) (b). The here presented data are smoothed and baseline corrected. The solid lines show the respective Gaussian fit.

The strong attenuation of the S2p signal indicates standing molecules with the sulfur bonded to the substrate [4]. The S2p peak of 11-MUA on Au(111)/mica (Fig. 2a) occurs at a binding energy of 162.7 eV, which corresponds to sulfur bonded to the gold substrate, in contrast to free sulfur (164 eV) and oxidized sulfur (168 eV) [4]. In Fig. 2b the S2p peak of 11-MUA on a recrystallized gold foil is displayed. Interestingly, one can notice a shift of the peak up to 163.3 eV. This higher binding energy might indicate the existence of partially non bonded sulfur. This corresponds to the observed larger S2p signal in this case, which shows that not all of the thiol head groups are attached to the surface. Atomic Force Microscopy. In general, it is quite difficult to detect a well ordered SAM on a smooth surface. However, defects or aggregates will be visible in AFM. In Fig. 3, a series of AFM images of a well prepared 11-MUA SAM on Au(111)/ mica are shown. The images were taken after different delay times under ambient conditions upon removal out of the solution. Figure 3a shows the surface directly after removal. A large island with a height of 6 nm can be observed (as determined from the cross section). With increasing delay time, 2 h (Figure 3b), the island shrinks considerably until after 4 h (Fig. 3c) the island has disappeared. Apparently, this

Fig. 3. 8 × 8 µm images of a 11-MUA film on Au(111)/mica as a function of delay time after removal out of the solution. (a) Immediately after removal (b) after two hours (c) after four hours

104

J. Stettner et al.

are aggregates of 11-MUA located above the SAM. According to ongoing investigations of PVD – preparations of 11-MUA SAMs at low temperature, we could find out that multilayers of 11-MUA desorb around room temperature [13]. Therefore we conclude that the observed island corresponds to 11-MUA multilayers. This relatively low desorption temperature of the multilayer can be seen as an advantage in comparison to other SAM systems, where long and careful ethanol rinsing is necessary to get rid of the multilayer [14]. Thermal Desorption Spectroscopy. TDS of 11-MUA was performed after installation of the samples into the UHV chamber and evacuation down to 10–8 mbar. Prior to these experiments, the cracking pattern of 11-MUA in the QMS was determined by direct evaporation from the Knudsen cell into the mass spectrometer. This was also corroborated by multilayer desorption of 11-MUA after PVD preparation [13]. In Fig. 4a the QMS cracking product mass 199 is shown for 11-MUA desorption (which corresponds to a cracking product of the intact molecule) from a gold foil and Au(111)/mica. For the smooth and densely packed (111) gold surface one single desorption peak develops, indicating desorption from the well ordered monolayer. Interestingly, desorption from the gold foil exhibits a second peak at higher temperature, which indicates differently bonded molecules. The reason might be that in this case flat lying molecules are present. In order to consider long term effects, SAMs of 11-MUA were kept under ambient conditions for four weeks. The results for subsequent TDS are shown in Fig. 4b. The total amount of mass 199 is not noticeably changed, but the desorption trace is quite different. In particular, the peak at 550 K is significantly decreased. Whereas desorption from the gold foil does not show any well pronounced feature anymore, desorption from Au(111)/mica yields two broad peaks. According to Horn et al. [15] a change in the tilt angle was observed for simple alkanethiols exposed to air over a period of several months using IR spectroscopy.

Fig. 4. Thermal desorption spectra for 11-MUA form Au(111)/mica and polycrystalline gold foil, respectively, shown for freshly prepared samples (a) as well as for samples which were kept under ambient conditions for one month (b).

Characterization of Self Assembled Monolayer Formation

105

Therefore, we suggest that the high temperature peak can be referred to molecules more strongly bonded to the substrate exhibiting a higher tilt angle, or again to flat lying molecules.

4.

Conclusions

11-MUA forms well ordered monolayers on the Au(111)/mica substrate, as shown by XPS and TDS. On the recrystallized gold foil, a less ordered MUA film is produced as shown by the changed binding energy of the sulfur measured by XPS revealing the presence of unbound sulfur. Thermal desorption of 11-MUA from Au(111)/mica yields only one desorption peak at a temperature of 550 K, whereas on the gold foil, a second peak shows up at higher temperature indicating differently bonded molecules. Prolonged exposure to air (four weeks) leads to considerable changes in the desorption spectrum. Acknowledgements. We would like to thank the FWF for the financial support (P19197 and S9702-N08).

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

A. Ulman, Chemical Reviews, 96, 1535, 1996. R. G. Nuzzo, L. H. Dubois and D. L. Allara, J. Am. Chem. Soc., 112, 558, 1984. C. Vericat, M. E. Vela and R. C. Salvarezza, PCCP, 7, 3258, 2005. S. Mendoza, I. Arfaoui, S. Zanarini, F. Paolucci and P. Rudolf, Langmuir, 23, 582, 2006. B. L. Frey and R. M. Corn, Analytical Chemistry, 68, 3187, 1996. P. Jiang, Z. Liu and S. Cai, Langmuir, 18, 4495, 2002. R. V. Duevel and R. M. Corn, Analytical Chemistry, 64, 337, 1992. K. Nakano, T. Yoshitake, Y. Yamashita and E. F. Bowden, Langmuir, 23, 6270, 2007. H. Wang, S. Chen, L. Li and S. Jiang, Langmuir, 21, 2633, 2005. E. Ito, K. Konno, J. Noh, K. Kanai, Y. Ouchi, K. Seki and M. Hara, Applied Surface Science, 244, 584, 2005. C. Gorman, Y. He and R. L. Carroll, Langmuir, 17, 5324, 2001. P. Frank, G. Hlawacek, O. Lengyel, A. Satka, C. Teichert, R. Resel, A. Winkler, Surface Science, 601, 2152, 2007. P. Frank, J. Stettner, F. Nußbacher and A. Winkler, in this proceedings. D. Käfer, G. Witte, P. Cyganik, A. Terfort and C. Wöll, J. Am. Chem. Soc., 128, 1723, 2006. A. B. Horn, D. A. Russell, L. J. Shorthouse and T. R. E. Simpson, J. Chem. Soc., 92, 4759, 1996.

SAMs of 11-MUA Grown on Polycrystalline Au-foils by Physical Vapor Deposition in UHV P. Frank, F. Nussbacher, J. Stettner and A. Winkler Institute of Solid State Physics, Graz University of Technology, Austria E-mail: [email protected] Abstract. In the present study, our aim was to prepare and characterize 11-mercaptoundecanoic acid (11-MUA) films on polycrystalline Au-foils grown by physical vapor deposition (PVD) in ultra high vacuum (UHV). When preparing the SAMs by PVD in a UHV chamber, one profits from well defined conditions and exclusion of the influence of air. The Au-samples were cleaned in UHV by Ar-sputtering. In order to characterize the 11-MUA films, mainly thermal desorption spectroscopy (TDS) was applied. TDS measurements revealed monolayer and multilayer desorption. A variety of different masses was observed when desorbing the 11-MUA from the Au-surface. This showed on the one hand the cracking pattern of this molecule in the quadrupole mass spectrometer (QMS) and on the other hand that reactions took place during the adsorption/desorption process. Of particular interest was the question as to the Au-S and the C-S bond breaking.

1.

Introduction

Self assembled monolayers (SAMs) consist of a single layer of ordered molecules on a substrate. These molecules consist of a functional end group, a spacer group (backbone) and a terminal anchoring unit. A SAM is formed due to the interplay of the anchor-substrate interaction and the lateral interaction of the molecules. SAMs have several applications in scientific research [1–3], which utilize in general the possibility to modify the surface properties and to perform structuring and contact printing, which can be used for novel organic devices. The desired SAM can be grown by immersion in solution or by physical vapor deposition (PVD) in ultra high vacuum (UHV). Most frequently, SAMs are grown by immersion in solution because the SAM preparation and the change of the functional end group (chemical tailoring) is easier to perform this way. The advantage of preparing the SAMs in UHV by PVD is obviously a cleaner substrate surface and the possibility to utilize surface science methods to characterize the grown SAMs. A frequently used prototype system is organothiols on gold. Apart from the customary functional end group, they consist of a terminal sulfur group and an organic backbone. There exists a large number of literature on SAM formation and characterization in general [1,3–5] and on functionalized alkane thiols in particular [6–8]. The most frequently used methods to analyze the SAMs are Scanning Tunneling Microscopy (STM), Infra-red spectroscopy (IRS), X-ray absorption spectroscopy (NEXAFS) and Low Energy Electron Diffraction (LEED). In this contribution we will focus on the application of Thermal Desorption Spectroscopy (TDS). This method allows

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_18, © Springer-Verlag Berlin Heidelberg 2009

107

108

P. Frank et al.

to get additional insight into the stability of self assembled monolayers and to investigate the adsorption/desorption kinetics. In particular, the reactivity of the functional end groups can be studied in detail by this method. For our studies we have used 11-mercaptoundecanoic acid (11-MUA) on gold surfaces as a model system.

2.

Experimental Methods

The MUA films on the Au foils were prepared under ultra-high vacuum (UHV) conditions by physical vapor deposition (PVD). The MUA molecules were deposited on the substrate by dosing from a heated stainless steel tube attached via a heated valve to a heatable glass container. Typically the stainless steel tube is held at 150°C and the reservoir at 50°C. As a substrate, we used a recrystallised gold foil because the main goal of this work was to compare our in-situ grown SAMs with SAMs grown ex-situ from solution [9], where similar substrates were used. Additionally, recrystallised gold foils offer the possibility to investigate the influence of the substrate geometry on the SAM formation, which has already been done successfully for other organic molecules [10]. The Au foil (10 mm × 10 mm, 0.1 mm thick) was attached to a steel plate (10 mm × 10 mm, 1 mm thick) by tantalum wires (0.25 mm in diameter). The steel plate itself was connected to the sample holder via tantalum wires and could be heated via these wires by resistive heating. The sample could be cooled down to 90 K by LN2 and heated to 1000 K. The temperature was measured on the backside of the steel plate by a Ni/NiCr thermocouple [11]. TDS was performed by desorbing the adsorbed molecules into a multiplexed line-of-sight quadrupole mass spectrometer (QMS). The heating rate for TDS was 1 K/s. After each TDS the substrate was cleaned by Ar-sputtering and subsequently annealed at 900 K. The cleanliness of the substrate after the cleaning process was checked by Auger electron spectroscopy (AES).

3.

Results and Discussion

Before the MUA adsorption experiments have been started one had to check the UHV compatibility, the cleanliness and the cracking pattern of the material. Due to the high vapor pressure at room temperature (TM = 50°C) evaporation from a Knudsen cell is not possible. As a consequence of this, thickness measurements with a quartz microbalance cannot be performed as well. The mass of the intact MUA molecule (HS-(CH2)10–COOH) is m = 218 amu, but due to the ionization process in the QMS most molecules are cracked. The result of this cracking process is a cracking pattern as shown in Fig. 1a. This pattern can only be obtained after prolonged degassing of the MUA material as received (Sigma Aldrich) at 400 K. During degassing significant water outgassing was observed. An alternative method to measure the cracking pattern of pure MUA is to directly desorb a thick MUA film into the QMS. In order to distinguish between cracking in the QMS and decomposition reactions taking place on the surface, the sample was cooled down to 200 K and a rather thick layer of MUA

SAMs of 11-MUA Grown on Polycrystalline Au-foils 109

was prepared. We can safely assume that the molecules which form the multilayer do not react with the substrate surface. A typical multilayer TDS featuring a clear zeroth order desorption is shown in Fig. 1b. In this case a higher source temperature (Tsource = 120°C) was used to get a high dosing flux. The cracking pattern of the multilayer shows the expected bunches of CxHx (e.g., m = 27 amu (C2H 3), m = 41 amu (C3H5), m = 55 amu (C4H7)), H2S (m = 34 amu) and COOH (m = 45 amu). Additionally, a small signal at the mass number of m = 218 amu (intact molecule) can be observed as well as a signal at m = 199/200 amu, which we assigned to the cracking of the acid end group of the intact molecule. For TDS, the QMS was tuned to the representative masses with the highest intensity of each species (e.g., mass 34, mass 41, mass 199).

Fig. 1. (a) Cracking pattern of MUA (b) TDS of multilayer regime, Tsource = 120°C

In Fig. 2a, the TDS of a MUA film after an evaporation time of 30 minutes (source temperature 50°C) at a sample temperature of 200 K is shown. This TD spectrum was taken directly after the MUA adsorption. It is important to mention this, because the characteristics of the monolayer regime depend on the time that passes between adsorption and desorption. Because of its small intensity, the signal of mass 199 was magnified by a factor of five in order to compare it with the other masses. There are two significant peaks, one of the multilayer regime (α, compare Fig. 1b, mind the different scaling) around room temperature and one of the monolayer regime around 525 K which actually consists of two peaks (β1 + β2). It’s remarkable that the cracking pattern of the monolayer (β2) is significantly different from the multilayer whereas the β1 peak exhibits a similar cracking pattern. This shows that at 510 K intact molecules of the SAM desorb but that the β2 peak originates from a decomposition product. The fact that no mass 199 can be seen but mass 34 (H2S) leads to the conclusion, that in this case the C-S bond is broken. This idea is supported by literature dealing with alkanethiols [12] and similar self-assembling thiols [13] on gold, as well as on other metals, for example on Cu [14,15]. A different TDS spectrum is obtained if the time between MUA film preparation and desorption is considerably increased (Fig. 2b). This film was prepared at 200 K and then left in the UHV for 115 hours. During that time the sample was not cooled and therefore a thawing of the sample took place, which led to the desorption of the multilayer. Again, the signal of mass 199 was magnified

110

P. Frank et al.

by a factor of five for comparison. In Fig. 2b, four peaks can be seen, which can be attributed to two different species. The peaks at 340 K, 440 K and 510 K (β1) show the same cracking pattern as the multilayer peak (only mass 41 and mass 199 are displayed), whereas the peak at 535 K (β2) contains the masses 41 and 34 but no mass 199. The peaks at 340 K and 440 K most probably stem from residual molecules of the multilayer. The first significant difference to Fig 2.a is that the total amount of the chemisorbed layer has more than doubled. We attribute this to the extended time available to arrange the monolayer in the presence of the multilayer during thawing. The second significant feature is that β1 and β2 can be better discriminated. This could be a hint to a better ordering of the SAM. It should be emphasized once more that the two states (β1 and β2) stem clearly from different species (HS-(CH2)10–COOH and (CH2)10–COOH). First comparisons with our ex-situ grown SAMs show similarities as well as discrepancies [9,16]: Of course, the multilayer-peak cannot be seen when the SAMs are prepared ex-situ at room temperature. Apart from that, the spectra are quite similar up to a temperature of about 650K, but around 700K an additional peak shows up, which is explained in more detail in Ref. [16].

Fig. 2. (a) TDS of a thick MUA film directly after adsorption at 200 K (b) TDS of an equivalent MUA film after thawing to room temperature and a waiting time of 115 h.

TDS also allows to determine the energetics and kinetics of adsorbates. From the multilayer desorption peak (Fig. 1b) we can obtain the desorption energy (heat of evaporation) and the pre-exponential factor (ν) by plotting lnR vs 1/T (Arrhenius plot): According to the Polanyi-Wigner equation [17], the gradient of the straight line yields the desorption energy and its Y-axis intercept yields the frequency factor. For the desorption energy we find 1.1 eV, which is in good agreement with values given in the literature for physisorbed alkanethiol with comparable chain length [7]. To get the pre-exponential factor the quantitative desorption rate is required. We obtained this by correlating the chemisorption peak (Fig. 2b) with the density of one monolayer (5 × 1014 molecules/cm²). Hence we get ν0 = 7.5 × 1017±1 s–1, which is significantly higher than the commonly assumed value of 1 × 1013 s–1 [13,18]. However, such high pre-exponential factors for alkanethiols have already been reported in the literature [19,20]. Further details on pre-exponential factors of large organic molecules can be found in Ref. [21–24]. We assume that the pre-exponential factor of the multilayer is in the same range as the one of the

SAMs of 11-MUA Grown on Polycrystalline Au-foils 111

monolayer and insert ν0 = ν1 = 7.5 × 1017±1 s–1 into the Redhead equation [17] for first order desorption. Hence we get Edes = 1.9 eV. The Tmax of the monolayer-peak of MUA on gold is similar to the values given in the literature for alkanethiols on gold [7,12,18,19]. The reason why our value for the desorption energy is somewhat higher than the ones given in the literature (usually from 1.2 eV to 1.5 eV) [7,18] is due to the higher value for the pre-exponential factor obtained in our case.

4.

Conclusions

The desorption kinetics of PVD grown MUA films on Au foils were investigated by thermal desorption spectroscopy. It was shown that the films grown at 200 K consist of a physisorbed multilayer (T des = 310–340 K) and a chemisorbed monolayer (Tdes = 500 K–550 K). Monolayer desorption actually proceeds in form of two peaks, one of the intact molecule and one of a fragment molecule lacking the thiol anchoring group. The desorption spectra of the chemisorbed layer changed with increasing waiting time giving information on the stability of the film. Acknowledgements. We acknowledge financial support by FWF, proj. no. 19197

References 1 2 3 4 5 6 7 8 9 10 11

12 13 14 15 16

A. Ulman, Chem. Rev., 96, 1533–1554, 1996. G. E. Poirier, Chem. Rev., 97, 1117–1127, 1997. F. Schreiber, J. Phys.: Condens. Matter, 16, R881-R900, 2004. J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo, G. M. Whitesides, Chem. Rev. 105, 1103, 2005. F. Schreiber, Prog. Surf. Sci. 65, 151, 2000. R. G. Nuzzo, L. H. Dubois, D. L. Allara, J. Am. Chem. Soc. 112, 558, 1990. F. Schreiber, A. Eberhardt, T. Y. B. Leung, P. Schwartz, S. M. Wetterer, D. J. Lavrich, L. Berman, P. Fenter, P. Eisenberger, G. Scoles, Phys. Rev. B 57, 12476, 1998. S. D. Evans, L. M. Williams, in Functional Organic and Polymeric Materials, Ed. T. H. Richardson, John Wiley Sons, 2000. J. Stettner, P. Frank, T. Griesser, G. Trimmel, R. Schennach, R. Resel and A. Winkler, in this proceedings. S. Muellegger, S. Mitsche, P. Pölt, K. Hänel, A. Birkner, C. Wöll and A. Winkler, Thin Solid Films 484, 408–414, 2005. P. Frank, G. Hlawacek, O. Lengyel, A. Satka, C. Teichert, R. Resel and A. Winkler, Surf. Sci., 601, 2152–2160, 2007. C. Kodama, T. Hayashi, H. Nozoye, Appl. Surf. Sci, 169–170, 264–267, 2001. D. Käfer, G. Witte, P. Cyganik, A. Terfort and C. Wöll, J. Am. Chem. Soc., 128, 1723–1732, 2006. A. Kühnle, S. Vollmer, T. R. Linderoth, G. Witte, C. Wöll and F. Besenbacher, Langmuir 18, 5558–5565, 2002. S. Vollmer, G. Witte and C. Wöll, Langmuir 17, 7560, 2001. J. Stettner, P. Frank, T. Griesser, G. Trimmel, R. Schennach, E. Gilli and A. Winkler, submitted to Langmuir.

112

P. Frank et al.

17 P. A. Redhead, Vacuum 12, 203, 1962. 18 D. J. Lavrich, S. M. Wetterer, S. L. Bernasek and G. Scoles, J. Phys. Chem. B, 102, 3456–3465, 1998. 19 H. Kondoh, C. Kodama, H. Sumida and H. Nozoye, J. Chem. Phys., 111, 1175, 1999. 20 S. L. Tait, Z. Dohnalek, C. T. Campbell and B. D. Kay, J. Chem. Phys., 122, 164708, 2005. 21 A. Winkler, in this proceedings. 22 K. A. Fichthorn and R. A. Miron, Phys. Rev. Lett., 89, 196103–1, 2002. 23 K. A. Fichthorn, K. E. Becker and R. A. Miron, Catalysis Today, 123, 71–76, 2007. 24 K. R. Paserba and A. J. Gellman, Phys. Rev. Lett., 86, 4338, 2001.

Photoreactive Self Assembled Monolayers for Tuning the Surface Polarity Thomas Griesser1,2*, Anna Track3, Georg Koller3, Michael Ramsey3 Wolfgang Kern1,2 and Gregor Trimmel 1* 1

Institute for Chemistry and Technology of Materials, Graz University of Technology, Stremayrgasse 16, Graz, Austria E-mail: [email protected] 2 Institute of Chemistry of Polymeric Materials, Montanuniversität Leoben, Franz-Josef Strasse 18, 8700 Leoben, Austria. E-mail: [email protected] 3 Institute of Physics, University of Graz, Universitätsplatz 5, 8010 Graz, Austria Abstract. In this contribution the modification of gold surfaces by self assembled monolayers (SAMs) of the photoreactive compound 11-mercaptoundecanoic acid, phenyl ester (MUAP) is presented. Upon irradiation with UV-light (λ = 254 nm) the phenyl ester groups photoisomerize to give hydroxyketones (photo-Fries reaction). Due to the formation of polar hydroxy groups the surface tension of the SAMs changes. The photogenerated hydroxyl groups were selectively modified with perfluorobutyryl chloride. This postexposure modification led to a significant change in wetting behaviour and surface energy and XPS indicates that patterning is possible.

1.

Introduction

The functionalization of inorganic surfaces by self assembled monolayers (SAMs) is a widely applied and important technique for the fabrication of nanostructured and hierarchically organised materials [1]. By the use of functional thiols the surface properties of inorganic substrates can be tuned. Depending on the functionalities employed, chemical reactions and immobilization of a broad variety of molecules are possible [2]. For many applications two dimensional patterning of surface properties and site-selective immobilization is required. A very convenient patterning technique is the use of photolithographic methods which allow the direct structuring of homogeneous SAM-films. In this context, SAMs bearing a photoreactive group are an interesting option. However, there are only a few photoreactions investigated on thiol-SAMs, e.g., the photocleavage of ortho-nitrobenzyl esters [3] and the photoisomerisation of azobenzenes [4]. In this contribution we present the synthesis of the new thiol, 11mercaptoundecanoic acid phenyl ester (MUAP), bearing photoreactive phenyl ester. Upon illumination with UV-light, the phenyl ester groups isomerize to the corresponding hydroxyketones (photo-Fries rearrangement [5]). We have recently shown, that this photoreaction in combination with subsequent post-modification reactions allows tuning the surface properties of polymeric layers [6]. Here, we investigate the modification of gold substrates by the chemisorption of MUAP, as well as by the photo-Fries reaction performed in the SAM layers. In addition, by

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_19, © Springer-Verlag Berlin Heidelberg 2009

113

114

T. Griesser et al.

immobilization of fluorocarbons on the illuminated surface, the formation of low energy surfaces was investigated by contact angle measurements and X-ray photoelectron spectroscopy.

2.

Experimental Methods

Synthesis of 11-Mercaptoundecanoic acid phenyl ester (MUAP). To a solution of 11-bromoundecanoic acid (10 g; 37.71 mmol) and phenol (3,23 g; 34,3 mmol) in 100 mL of CH2Cl2 10.62 g (51.47 mmol) of N,N-dicyclohexyl-carbodiimid were added. The mixture was stirred at room temperature for 20 h. The reaction was quenched with aqueous HCl (5%, 20 mL) and stirred for 5 min. The reaction mixture was filtered over Celite. The filtrate was washed with saturated sodium bicarbonate solution (50 mL) and water (3 × 50 mL). After evaporation of the solvent the crude product was purified by column chromatography (ethylacetate: cyclohexane 1:20) to give 8.67 g (67.36%) of 11-bromoundecanoic acid phenyl ester (BrUAP). A solution of BrUAP (5.0 g; 14.65 mmol) in 30 mL of tetrahydrofurane (THF) was cooled to –10°C and hexamethyl-disilathiane (3.14 g; 17.6 mmol) and tetrabutyl ammonium fluoride (5.08 g; 16.1 mmol in 17 ml THF with 170 µl water) were added. The reaction mixture was stirred at room temperature for 1 h and was then diluted with 60 mL of CH2Cl2 and washed with aqueous ammonium chloride (3 × 30 mL). After evaporation of the solvent the crude product was purified by column chromatography (ethyl acetate:cyclohexane 1:20) to give 2.05 g (47.52%) of MUAP. 1 H-NMR: (δ, 500 MHz, 20°C, CDCl3): 7,38 (t, 2H, ph3,5); 7,22 (t, 1H, ph4); 7,07 (d, 2H, ph2,6); 2,57–2,50 (4H, C2,11); 1,78–1,72 (m, 2H, C3); 1,64–1,58 (m, 2H, C10);1,43–1,29 (m, 12H, C4–9) ppm. 13C-NMR: (δ, 125 MHz, 20°C, CDCl3): 172,28 (1C, C = O), 150,70 (1C, ph1), 129,36 (2C, ph3,5), 125,68 (1C, ph4), 121,54 (2C, ph2,6), 34,37–24,64 (10C, C2–11) ppm. IR (CaF2, cm–1): 2929; 2855; 1760; 1594; 1493; 1465; 1457; 1364; 1197; 1163; 1141; 1070; 1024; 929. Preparation of SAMs. Freshly cleaned gold substrates (Arrande, Germany) were immersed into a 2 mM solution of MUAP in ethanol. After 24 h the substrates were rinsed with ethanol and dried in a nitrogen stream. UV-Illumination. was carried out in inert atmosphere (argon) by using an ozone free mercury low pressure UV lamp (Heraeus Noblelight; 254 nm) with a power density of 1.35 mW/cm2. One half of the substrate was shaded during illumination. Post-modification Reaction. After the illumination step, the substrate was immersed in a solution of 200 µL perfluorobutyryl chloride and 50 µL triethyl amine in 2 mL CH2Cl2 for 20 min. The substrates were washed with CH2Cl2 and dried in a stream of nitrogen. Contact Angle Measurements. were obtained with a Drop Shape Analysis System DSA100 (Krüss GmbH, Hamburg, Germany) using water and diiodomethane as test liquids. The contact angles were measured by the sessile drop method within two seconds. The surface tension γ as well as the dispersive and polar components (γD and γP) were calculated based on the Owens-Wendt method [7].

Photoreactive Self Assembled Monolayers 115

XPS-measuerements were carried out with an XPS equipment (SPECS) in ultra high vacuum (1 × 10–10 mbar) using MgKα radiation (1253.6 eV) and a hemispherical analyzer (Phoibos100) with an energy resolution of 1.2 eV. All spectra were normalised to the Au4f7/2-peak.

3.

Results and Discussion

The NMR and IR-spectroscopic data of the new photoreactive thiol MUAP are in good agreement with the proposed structure. SAMs were prepared by immersing gold substrates into a solution of MUAP in ethanol for 24 hours. In the next step, only one half of these films was illuminated using UV-light of 254 nm. Thereby the phenyl ester groups undergo the photo-Fries rearrangement to the corresponding hydroxyketones in the illuminated areas, as schematically depicted in Fig. 1. The photogenerated hydroxy-groups are more reactive than the ester groups as demonstrated by a successive post-modification step with perfluorobutyryl chloride. The fluorinated compound can only be immobilized on the illuminated side.

O

HO

O

SAM-

FF F F F F O F O O

O

CF3(CF2)2COCl

UV-light

Pyridine

formation

Au

S

S

S

Au

Au

Au

water

diiodomethane Fig. 1. Scheme for the SAM formation of MUAP, the photo-Fries rearrangement reaction upon irradiation with UV-light, and the post-modification reaction with perfluorobutyryl chloride. Below: images of water droplets and CH2I2 droplets on each surface.

116

T. Griesser et al.

This reaction scheme was investigated by contact angle measurements and XPS. Images of water and diiodomethane droplets of the original gold substrate, of the SAM before and after UV-illumination, as well as after the post-treatment with perfluorobutyryl chloride are presented in Fig. 1. The contact angles and the calculated surface tension are summarized in Table 1. Table 1. Contact angle (θ, sessile drop) and surface tension data (γ: surface tension; γD: dispersive component, γP: polar component; surface polarity = 100 (γP/γ) of the gold substrate, the SAM layer before and after illumination, and with subsequent post-modification with perfluorobutyryl chloride.

gold MUAP MUAP illuminated MUAPilluminated + modified

θ H O2 / ° 88 ±3 91 ±1 78 ±1

θ CH2I2 / ° γ/mJ/m2

γD/mJ/m2

γP/mJ/m2 polarity / %

23 ±1 32 ±3 35 ±0

47.5 ±0.2 43.8 ±0 45.8 ±0.1

46.8 ±0.1 43.4 ±0 42.2 ±0

0.7 ±0.1 0.4 ±0 3.7 ±0

1.5 ±0.1 0.9 ±0 8.1 ±0.1

93 ±1

52 ±1

34.2 ±0.1

33.1 ±0.1

1.1 ±0

3.2 ±0.1

Upon formation of the SAM the contact angles increase both for water (from 88° to 91°) and for CH2I2 (from 23° to 32°) which is the result of a reduced overall surface tension caused by the chemisorption of the MUAP. Upon UV-illumination the contact angle of water decreases to 78° whereas the contact angle of CH2I2 only slightly increases to 35°. The surface tension is slightly enhanced, however, the polarity has the highest value of the investigated samples (8.1%). This can be attributed to the presence of the polar hydroxy groups formed by the photoreaction. After treatment with perfluorobutyryl chloride, both contact angles rise to highest observed values of 94° and 52°, respectively. The surface tension is with a value of only 34.2 mJ/m2 the lowest value of this series due to the immobilisation of fluorocarbons. XPS-spectra confirm the SAM formation as the corresponding carbon, sulfur and oxygen signals were found. The XPS-spectra hardly change due illuminated half non-illuminated half

680

685

690

695

binding energy / eV

Fig. 2. Comparison of the F1s region of the XPS spectra after treatment with perfluorobutyryl chloride of the non-illuminated side and the illuminated side.

Photoreactive Self Assembled Monolayers 117

to the photo-Fries reaction, as no change of the chemical composition occurs. However, after the post-modification reaction with perfluorobutyryl chloride, only the illuminated side shows an intense signal in the F1s region (Fig. 2). This clearly indicates that selective surface chemistry is occurring and patterning is possible.

4.

Conclusions

In this contribution we have presented the synthesis of the new photoreactive thiol MUAP bearing a photoreactive phenyl ester. This molecule is well suited for the preparation of SAMs on gold substrates and can be photochemically modified by UV-illumination due to the photo-Fries rearrangement leading to the formation of hydroxyketones. By a subsequent post-modification reaction with a fluorinated compound low-energy surfaces are attainable. Acknowledgements. Financial support by the FWF within the NFN “Interface Controlled and Functionalised Thin Organic Films“ (S9702-N08 and S9704-N08) is gratefully acknowledged.

References 1 2 3 4 5

6 7

J.C. Love, L.A. Estroff, J.K. Kriebel, R.G. Nuzzo, G.M. Whitesides in Chemical Reviews 105, 1103, 2005. X. Li, J. Huskens, D.N. Reinhoudt in Journal of Materials Chemistry 14, 2954, 2004. K. Critchley, J.P. Jeyadevan, H. Fukushima, M. Ishida, T. Shimoda, R.J. Bushby, S.D. Evans in Langmuir 21, 4554, 2005. A. Manna, P.-L. Chen, H. Akiyama, T.-X Wie, K. Tamada, W. Knoll in Chemistry of Materials 15, 20, 2003. M.A. Miranda and F. Galindo, Photo-Fries Reaction and Related Processes in CRC Handbook of Organic Photochemistry and Photobiology, 2nd ed.; W.M. Horspool, Ed.; CRC Press: Boca Raton, FL, 2004. T. Griesser, T. Hoefler, S. Temmel, W. Kern, G. Trimmel in Chemistry of Materials 19, 3011, 2007. D.K. Owens and R.C.J. Wendt in Journal of Applied Polymer Science 13, 1741, 1969.

Spectroscopy of Defects in Epitaxially Grown Para-sexiphenyl Nanostructures A. Kadashchuk 1,2, S. Schols2, Yu. Skryshevski1, I. Beynik3, C. Teichert 3, G. Hernandez-Sosa4, H. Sitter4, A. Andreev5, P. Frank6 and A. Winkler6 1

Institute of Physics, Natl. Academy of Sci. of Ukraine, Kiev 03028, Ukraine E-mail: [email protected] 2 IMEC v.z.w., SOLO-PME, Leuven B-3001, Belgium 3 Institute of Physics, Montanuniversitaet Leoben, Leoben A-8700, Austria, 4 University of Linz, Linz A-4020, Austria 5 Nanoindent Technologies AG, Linz A-4040, Austria 6 Graz University of Technology, Graz A-8010, Austria Abstract. We present a study of steady-state- and time-resolved photoluminescence of para-sexiphenyl (PSP) films on KCl grown by organic molecular beam epitaxy (OMBE). Using different OMBE growth conditions has enabled us to vary greatly the morphology of the PSP crystallites but keeping virtually untouched their chemical structure. By this comparative study we prove that the broad red-shifted emission band has a structure-related origin rather than being due to monomolecular oxidative defects. The relative intensity of the defect emission band observed in the delayed spectra was found to be drastically suppressed in the OMBE-grown films dominated by growth mounds composed of upright standing molecules as opposed to the films consisting of crystallites formed by molecules lying parallel to the substrate.

1.

Introduction

Para-sexiphenyl (PSP) is a very attractive material for the application in blue OLEDs [1] due to its high quantum efficiency and also for future electrical-pumping organic solid-state lasers as it combines a high electroluminescent efficiency and high charge-carrier mobility, along with a low lasing threshold [2]. A special advantage of PSP for device fabrication lies in its ability to self-organize in well-ordered crystallites at the macroscopic scale under organic molecular beam epitaxy (OMBE) or hot wall epitaxy (HWE) conditions on underlying mica or alkali halides. Recently a low-energy broad emission band centered around 480 nm was found [3, 4] in highly ordered PSP films grown by HWE on mica, while the intrinsic exciton emission of the film is in the blue region (0-0 transition is at ~400 nm). It should be mentioned that the 400nm peak is subject to strong reabsorbed in thick PSP layers and 0–1 transition at ~420 nm can dominate the PL spectra in such case. The origin of the observed red-shifted emission was attributed to interchain excitations [4] due to some sort aggregate species, i.e. structure defects in the PSP films grown on mica. However, a rather similar phenomenon was found in some conjugated polymers (as polyfluorenes (PF) and poly(p-phenylenes) (PPP)) and has been interpreted in terms of either aggregate/excimers formation [5] or due to

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_20, © Springer-Verlag Berlin Heidelberg 2009

121

122

A. Kadashchuk et al.

the presence of ketonic defects (segment carrying a fluorenone oxidation product) on the main chain [6]. In this paper we focus on time-resolved photoluminescence (PL) studies of PSP films grown by OMBE on KCl. Using different OMBE growth conditions enabled us to vary greatly the morphology of the PSP crystallites but keeping their chemical structure virtually untouched. We show that the red-shifted emission band has clearly a structure-related origin rather than being due to monomolecular oxidative defects.

2. Experimental Methods PSP films with a mean thickness of 30 nm have been grown by OMBE techniques on freshly cleaved (001)-oriented KCl(001) substrates at different substrate temperatures, Tsub, ranging from 250 K to 450 K under ultra-high vacuum condition. Further details on the film growth and the surface characterization can be found elsewhere [7]. The film morphology was imaged ex-situ by Atomic Force Microscopy (AFM). Measurements of delayed emission were carried out using a laser pulse excitation at 337 nm and a gating registration system based on an intensified CCD camera. A variable delay from 75 ns to 10 ms after optical excitation allowed the detection of weak delayed luminescence after the intense prompt fluorescence.

3. Results and Discussion Figure 1 shows AFM images of the surface morphology of PSP films grown by OMBE on crystalline KCl substrates at Tsub 250 K (a, c) and 450 K (b, d), respectively. To reveal three-dimensional information, line scans through the 1 µm × 1 µm images are also presented as insets. This AFM results demonstrate a strong influence of the substrate temperature on the growth morphology. At Tsub = 250 K the film is dominated of small, rather irregularly arranged features with lateral dimensions between 30 to 60 nm in width and 50 to 100 nm in length, with heights up to 20 nm. At Tsub of 300 K, we observe that the structures elongate along the <110> directions of the KCl(001) substrate up to lengths of 500 nm (not shown here), therefore we can assume that the small features at 250 K are precursors to the needle-like features formed at room temperature. From x-ray data it is well known that these needle-like structures are composed of molecules that are parallel to the substrate [8]. With increasing Tsub, growth mounds composed of terraces with 2.6 nm height (inset in Fig. 1b) become progressively dominant, though a few needle structures remain visible. From the terrace height we can conclude that these mounds are composed of molecules that are almost upright standing on the substrate as was supported by x-ray investigations [8].

Spectroscopy of Defects in Epitaxially Grown Para 250K a)

nm

1 µm

100

450K b)

<110>KCl

nm

50

1 µm

<110>KCl

<110>KCl

<110>KCl 0

c) 20

0 d) 10 8 6 4 2 0 z [nm]

z [nm]

123

0 0

0.1

0.2 0.3 x [µm]

0.4

0

0.2

0.4 x [µm]

0.6

Fig. 1. AFM images (5 µm × 5 µm ) of the surface morphology of 30 nm PSP films grown by OMBE on KCl at two different substrate temperatures: 250 K (a) and 450 K (b). The insets show corresponding 1 µm × 1 µm AFM images. The corresponding cross-sections along the white lines indicated in the insets are presented in (c) and (d).

Temperature-dependent cw-PL spectra from the above PSP films grown at 250 K and 450 K are shown in Fig. 2a and 2b, respectively. These spectra exhibit a characteristic structure in the blue spectral region as reported before for HWEgrown PSP nanocrystallites [4]. Note that the cw-PL spectra of both standing (Tsub = 450 K) and lying (Tsub = 250 K) films do not reveal any apparent broad band at 480 nm (Fig. 2) with increasing temperature in contrast to the HWE-grown films on mica [4]. Figure 3 shows room-temperature delayed emission spectra from PSP films grown at 450 K and 250 K (curve 1 and 2, respectively) measured under the same conditions. Remarkably that the delay emission spectra of these OMBE-grown films of different morphologies (Fig. 1) are drastically different; namely, the film grown at 450 K is clearly dominated by the intrinsic delayed fluorescence with a maximum at ~420 nm, while the low-energy defect band centered at ~480 nm dominates in the film grown at 250 K. The results presented above show that the low-energy broad PL band at 480 nm in the OMBE-grown films is seen only in the time-delayed emission spectra and it stems from some defects in the material and its relative intensity depends on the sample morphology. This defect band can in principle be either due to structural defects in epitaxially-grown nanocrystallites or related to some sort of “chemical” defects, as fluorenone defects possibly created by oxidation of PSP molecules. In contrast to the HWE grown films, generation of oxidative defects might be excluded in the course of film growth by the OMBE technique since it occurs under UHV conditions. Moreover, the above mentioned defect band is clearly seen in the delayed spectra of the lying film but not in the standing one (Fig. 3).

124

A. Kadashchuk et al.

6

cw-PL intensity, (arb. units)

4x10

a)

b)

Tsub=250K (lying molecules)

Tsub=450K (standing molecules)

1)= 5 K 2)= 80 K 3)=100 K 4)=125 K 5)=150 K 6)=175 K 7)=200 K 8)=225 K 9)=250 K 10)=275 K 11)=293 K

6

3x10

6

2x10

1)= 5 K 2)= 80 K 3)=100 K 4)=125 K 5)=150 K 6)=175 K 7)=200 K 8)=225 K 9)=250 K 10)=275 K 11)=293 K

5K 5K 6

1x10

293 K

0 400

450

Wavelength, (nm)

293 K

500

400

450

500

Wavelength, (nm)

Fig. 2. Temperature-dependent cw-PL spectra from PSP films grown by OMBE on KCl substrates at 250 K a) and 450 K b). Film thicknesses are the same and equal to 30 nm.

Fig. 3. Room-temperature delay PL spectra from PSP films grown by OMBE on KCl substrate at 450 K and 250 K (curves 1 and 2, respectively) using time delay of 75 ns and registration time gate width of 100 μs.

This observation provides a straightforward evidence for the structural origin of the broad low-energy band at 480 nm in the PSP films and eliminates its assignment to possible monomolecular oxidative defects. Finally it should be noted that the observed defect band at 480 nm, being shifted towards low energy by about 0.5 eV with respect to the 0-0 exciton transition, cannot be due to different HOMO level positions in planar (standing film) and twisted (lying film) conformation of the PSP molecules [9]. The similarity of the cw-PL spectra of the above films (Fig. 2) implies that the lowest excited singlet state position is almost the same for both film morphologies.

Spectroscopy of Defects in Epitaxially Grown Para

4.

125

Conclusion

We found that OMBE-grown PSP films on KCl substrates at different substrate temperatures show a drastic difference in the delay emission spectra while their cw-PL spectra are rather similar. The broad structureless defect emission band dominates the delayed PL emission of PSP films consisting of lying molecules on the substrate and no such band has been observed in films composed of upright standing molecules. This clearly indicates a structure-related origin of the observed defects, implying that their concentration could be minimized in most perfect structures of PSP crystallites. Acknowledgment. Research was implemented within the ÖAD Project UA 01/2007 and supported by the program of fundamental research of the National Academy of Sci. of Ukraine, “Nanostructured systems, nanomaterials, nanotechnologies” through the Project No. 10/07-H, and by the Ministry of Education and Science of Ukraine through Project No. M/138-2007. The support by the Austrian Science Fund (FWF), Project numbers P19197, P19636-N20 and S 9707-N08 is also gladly acknowledged.

References 1. 2. 3. 4. 5. 6. 7. 8. 9.

S. Tasch, C.Brandstätter , F. Meghdadi, G. Leising, et al. Adv. Mater., 9, 33, 1997. A. Andreev, et al., J. Appl. Phys. 99, 034305, 2006. A. Andreev, et al., Adv. Mater. 12, 629, 2000. A. Kadashchuk, A. Andreev, et al., Advan. Func. Mater. 14, 970, 2004. A. Haugeneder, U. Lemmer, U. Scherf, Chem. Phys. Lett. 351, 354, 2002. U. Scherf, E. List, Adv. Mater. 14, 477, 2002. P. Frank and A. Winkler, Appl. Phys. A 90, 717, 2008. T. Haber, M. Oehzelt, et al. J. Nanoscience and Nanotechn., 6, 698, 2006. M. Ramsey, et al. Adv. Mat. 15, 1812, 2003.

Magnetoresistance in Poly (3-hexyl thiophene) Based Diodes and Bulk Heterojunction Solar Cells S. Majumdar1,2, H. S. Majumdar1, H. Aarnio1, R. Laiho2 and R. Österbacka1 1

Center for Functional Material and Department of Physics, Åbo Akademi University, 20500 Turku, Finland. E-mail: [email protected] 2 Wihuri Physical Laboratory, Department of Physics, University of Turku. 20014 Turku, Finland. Abstract. We report on magnetotransport studies of regio regular poly (3-hexyl thiophene) (RRP3HT) based diodes and P3HT: 1-(3-methoxycarbonyl)propyl-1-phenyl-[6,6]methanofullerene (PCBM) bulk heterojunction solar cells. While the P3HT diodes with non-magnetic electrodes ITO and Al show up to 16% positive magnetoresistance (MR) under 300 mT magnetic field at room temperature, solar cells with much reduced electronhole pair (e-h) formation probability show almost negligible MR clearly suggesting that MR in organic semiconductors is strongly dependent on the probability of forming Coulombically bound e-h pairs. Voltage and temperature dependence of MR sign and magnitude is discussed for the diode devices which provides important insight of the scientifically puzzling yet technologically very promising magnetotransport properties of organic semiconductor based devices.

1.

Introduction

Organic Magnetoresistance (OMAR) for sensor and other device applications has gained impetus over the last few years since its first report in 2003 [1–4]. The uniqueness of OMAR is that this effect is observed in organic semiconductor (OS) thin film devices with no ferromagnetic electrodes – unlike the organic spin-valves [5]. The OS have very weak or no magnetic properties by themselves, however, as a device they exhibit sizable magnetoresistance (MR) at room temperature as well as low temperatures. So far the OMAR has been reported to be a very generalized observation in sandwich structures made from OS materials having hydrogen atoms in their side chains and hence showing hyperfine interaction. However, the physical mechanism of this effect is still not completely understood. Different models have been proposed so far, namely magnetic field induced singlet-triplet interconversion (MIST) model [3], triplet polaron quenching model [4] and bipolaron model [6], to explain the main features of OMAR but there is still some mismatch between experimental data and theoretical understanding. The generality of OMAR in different OS materials has been the advantage and still the least understood part of it. The independence of materials makes the effect very attractive in terms of applications. However, different sign of OMAR depending on different materials or operating conditions like bias, temperature etc. make it very

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_21, © Springer-Verlag Berlin Heidelberg 2009

127

128

S. Majumdar et al.

intriguing. The values of the %MR reported in literature also vary quite much [2,4] and there is often a maxima in the MR curve as a function of bias [4]. In this communication, we have studied the effect of magnetic field on charge transport dynamics in two systems, i.e. (i) in regio-regular poly (3- hexyl thiophene) (RRP3HT) diodes and in (ii) RRP3HT:PCBM bulk heterojunction solar cells (BHSC), where the probability of forming Coulombically bound e-h pair are reduced. Comparative study of these two systems will provide information about the effect of magnetic field on e-h recombination and explain the possible cause of OMAR.

2.

Experimental Methods

The device structure used in the present experiment is ITO/PEDOT:PSS/ RRP3HT/LiF/Al. The ITO coated glass electrodes were coated with a very thin layer of PEDOT:PSS and annealed at 120 deg for 15 minutes. The pi-conjugated polymer RRP3HT was spin-coated from a chloroform solution and annealed at 120 deg for 15 minutes. The approximate thickness of the RRP3HT layer was 150 nm. Finally the lithium fluoride and the aluminium electrode was vacuum evaporated to complete the device structure. For solar cells, a 1:1 blend of RRP3HT and PCBM was dissolved in dichlorobenzene and spin coated on PEDOT:PSS coated ITO/glass substrates with thickness of 150 nm and annealed and finally LiF and Al were evaporated. The device preparation was done in a nitrogen-filled glovebox using anhydrous solutions. After fabrication, the devices are transferred in nitrogen atmosphere to a cryostat placed between the pole pieces of an electromagnet capable of producing up to 300 mT magnetic field. The resistance of the device is then measured by sending a constant current through the device and measuring the voltage drop across it in varying magnetic field in the temperature range 100–300 K. For characterizing the solar cells, integral mode of time of flight, charge extraction by linearly increasing voltage and double injection transients were used which is described elsewhere [7].

3.

Results and Discussion

The %MR is defined as %MR = (RB – R0)/R0 × 100%, where RB and R0 is the device resistance in presence or absence of a magnetic field B. In the RRP3HT based diode devices, the %MR at room temperature shows positive value in most of the devices, whereas in the devices with higher reverse bias (leakage) currents a sign inversion of MR is observed depending on the bias currents (Fig. 1). For some devices, a maximum of up to 16% positive MR was also observed at room temperature. By changing the minority carrier injection with substitution of Al electrode by gold, this MR transition voltage can be modified. Hence, depending on the minority carrier injection the MR response of the device can be tuned between positive and negative values in these P3HT based diode devices.

Magnetoresistance in Poly (3-hexyl thiophene) Based Diodes

129

20 15

300 K

100 μA

% MR

10 5

10 μA 1 μA

0

100 nA 50 nA 20 nA

-5 -10

50

100

150

200

250

300

350

B (mT)

Fig. 1. % MR as a function of magnetic field for a typical RRP3HT based diode device with high leakage current at room temperature showing negative to positive switchover at a particular bias voltage.

We find that the MR response of the RRP3HT diode devices either starts to increase rapidly or change sign from negative to positive value for devices with higher leakage currents at a particular transition voltage (Vtr) which closely corresponds to the threshold voltage (Vth) of the device. The diodes, even the ones which do not show any transition from positive to negative MR at 300 K, start showing transition as we lower the temperature, however the one to one correspondence between Vth and Vtr is maintained at all the temperatures (Fig. 2 inset). Earlier, similar observation was reported by Bloom et al. [8] in Alq3 based diodes. With decreasing temperature Vtr starts increasing. However, below 150 K the device current decreases appreciably and MR response of the device can not be seen clearly anymore. Additionally we have observed that the MR line shape is a strong function of bias currents, temperatures, materials and device electrodes. In the bipolaron model [6], it is claimed that OMAR line shape is universal with either a Lorentzian B2/(B02 + B2) or a specific non-Lorentzian B2/(|B| + B0)2 fit, where B0 is the full width at half maximum of the MR traces and is seen for all materials and bias currents. It has been observed repeatedly for many devices that at lower bias currents the saturation magnetic field in the MR curve lies between 10–50 mT whereas with higher bias currents % MR increases more rapidly and saturation is not reached until 300 mT, clearly showing that the saturation field is a function of bias as well. The devices also have a strong tendency to retain its magnetic history for a considerable period of time and does not come back to its pristine state immediately after removing the magnetic field. However, detailed study of the bias dependent line shape and hysteresis phenomenon is out of scope of this paper and will be presented in a later communication. Next, we studied the magneto-transport properties on the conjugated polymer/ fullerene BHSC. In a BHSC, a polymer – fullerene blend is used, where fullerenes having higher electron affinity attract electrons in it and holes are left on the polymer chains. Thus, electron and hole pathways are separated givinga lower probability of e-h pairs to be formed.

130

S. Majumdar et al. 3 V th (V)

3.0

VTr (V)

2.5

2.0

Slope = 1

2

1 1

2

3 V tr(V)

1.5 140 160 180 200 220 240 260 280 300 320 340

Temperature (K)

Fig. 2. The negative to positive switchover voltage (Vtr) as a function temperature for a typical RRP3HT based diode device. Inset shows one to one correspondence of the diode threshold voltage (Vth) and Vtr at all the temperatures. P3HT Diode Solar cell

300 K

% MR

15

10

5

0 -300

-200

-100

0

100

200

300

B (mT)

Fig. 3. %MR as a function of magnetic field for a typical RRP3HT based diode and a P3HT:PCBM bulk heterojunction solar cell at room temperature showing almost negligible MR of good solar cells compared to the diodes.

We have used RRP3HT/PCBM blend where the bimolecular recombination is greatly reduced [9] and studied the effect of magnetic field on charge transport. Figure 3 shows the comparison between the MR response of a typical diode and a solar cell device. It is clearly seen that in a BHSCs with well-separated electron and hole channels, MR is orders of magnitude less than that of a diode. Now, in a BHSC, the exciton formation is minimized but the single carrier density inside polymer and fullerene is expected to be higher, resulting in much higher probablity of bipolaron formation. This result clearly indicates that magnetic field mainly affects the e-h paired state or excitonic state (when the e-h pair is captured inside a single molecule) and does not affect the single carrier movement so much as proposed in the bipolaron model for OMAR [7].

4.

Conclusions

In conclusion, we have observed both positive and negative magnetoresistance in RRP3HT based diode devices. The voltage, at which the MR sign change occurs is closely linked to the diode threshold voltge and a strong function of temperature.

Magnetoresistance in Poly (3-hexyl thiophene) Based Diodes

131

In a solar cell device with well-separated electron and hole channels, the MR value reduces dramatically showing clearly that the bipolaronic picture can not account for the magnetic field dependent transport properties in organic semiconductor based devices. Acknowledgements. The authors would like to thank financial support from Academy of Finland and Wihuri Foundation. S.M. also acknowledges Sitra- CIMO fellowship for carrying out part of this research work.

References 1 2 3 4 5 6 7 8 9

T. L. Francis, Ö. Mermer, G. Veeraraghavan and M. Wohlgenannt, New J. Phys., Vol. 6, 185, 2004. Ö. Mermer, G. Veeraraghavan, T. L. Francis, Y. Sheng, D. T. Nyugen, M. Wohlgenannt, A. Köhler, M. K. Al-Suti and M. S. Khan, Phys. Rev. B, Vol. 72, 205202, 2005. V. Prigodin, J. Bergson, D. Lincoln and A. Epstein, Synth. Met., Vol. 156, 757, 2006. P. Desai, P. Shakya, T. Kreouzis, W. P. Gillin, N. A. Morley and M. R. J. Gibbs, Phys. Rev. B, Vol. 75, 094423, 2007. S. Majumdar, H. S. Majumdar, P. Laukkanen, J. Värynen, R. Laiho and R. Österbacka, Appl. Phys. Lett, Vol. 89, 122114, 2006. P. A. Bobbert, T. D. Nyugen, F. W. A. van Oost, B. Koopmans and M. Wohlgenannt, Phys. Rev. Lett. Vol. 99, 216801, 2007. S. Majumdar, H. S. Majumdar, H. Aarnio, D. Vanderzande, R. Laiho and R. Österbacka, Phys. Rev. Lett. (submitted), arXiv: 0805.2546. F. L. Bloom, W. Wagemans, M. Kemerink and B. Koopmans, Phys Rev. Lett, Vol. 89, 122114, 2006. A. Pivrikas, N. S. Sariciftci, G. Juška and R. Österbacka, Prog. Photovolt: Res. Appl. Vol. 15, 677 (2007).

Evolution of the Bipolaron Structure in Oligo-diacetylene Films: A Semiempirical Study Massimo Ottonelli1, Gianfranco Musso, and Giovanna Dellepiane INSTM and Università di Genova, Dipartimento di Chimica e Chimica Industriale, via Dodecaneso 31, 16146 Genova, Italy 1 E-mail: [email protected] Abstract. The formation of polaron and/or bipolaron species upon chemical doping and photo- or electro-chemical generation is a scientific and technological problem which is still object of discussion. Here we use a simple electrostatic model to study the role of the interchain interaction in stabilizing or destabilizing the bipolaronic structure, which is usually neglected in theoretical studies. The results show that, as a consequence of the interchain interactions in the bulk system, a bipolaron generated on a single oligomer chain more than 12 units long will soon split into two polarons on adjacent chains, this process being maximized for specific interchain distances.

1.

Introduction

Conjugated organic polymers, or oligomers, are materials of great promise for their applications in the electronic technology [1]. In fact, it is well know that through the design of new structures or functionalities [2] the conductivity properties of this class of materials can be modulated over a wide range of values starting from those typical of an insulator until, after doping, to those of a metal [3]. Different from what happens in traditional inorganic semiconductors, in this matter a critical point are the effects induced in the conjugated material by charge injection. For polymer/oligomer systems the hole or electron injection induces backbone geometry distortions that act as confining potential for the excess charge itself. Such species are called a polaron or bipolaron if one or two charges respectively are introduced in the molecule. Knowledge of the relative stability of the polaron/bipolaron species in this class of materials is of great importance for applications, and still object of an open debate in the scientific community. In fact, the two-peak features observed in doped polymers [4] were at first assigned to the presence of bipolarons, a result that was further supported by theoretical calculations [5] and by subsequent experimental photoinduced absorption spectra of polythiophenes [6]. On the contrary recent experimental [7] and theoretical [8] studies have attributed the above features to the presence of polarons (or polaron pairs) rather than to bipolarons. Here we will contribute to this debate by introducing the effects of the interchain interactions, usually not taken into account in theoretical studies. Through a simple electrostatic model, based on semiempirical calculations, we will show that interchain interactions open a charge transfer channel between neighbour oligomers, so that a double (positive) charge injected on a single chain can fast split into two polarons on adjacent chains. This process results to be maximized in a specific range of interchain separations. H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_22, © Springer-Verlag Berlin Heidelberg 2009

133

134

M. Ottonelli, G. Musso and G. Dellepiane

2.

Results and Discussion

To describe one-dimensional conjugated systems, we choose as a model unsubstituted oligodiacetylenes (nPDAs, where n is the number of repeat units), since this class of materials is of potential interest for technological applications [9]. We focus our study on 3PDAs and 15PDAs. The former represents the typical case of a “confined” bipolaron since the short oligomer size does not allow a complete geometry relaxation. On the other hand 15PDAs is the shortest oligomer for which the self-induced geometry distortions due to the injection of a double (positive) charge find enough room to fully occur, and is sufficiently large to model the polymer properties. The absence of strong confinement effects in large oligomers has been shown by us [10] for oligodiacetylenes, and by other authors [11] for oligothiophenes. For shorter oligomers (below 12 repeat units) the geometry distortions induced in the backbone are those belonging to a classical bipolaron picture, while for 12 or more repeat units the excess charge splits into a polaron pair.

Fig. 1. AM1 total energy for the isolated neutral, single and doubly charged nPDAs (n = 3,15) as computed at the optimized geometry of the neutral or doubly charged oligomer. The charge of the oligomer is shown in brackets.

The total AM1 energies for three differently oxidized forms of the 3PDAs and 15PDAs isolated oligomers are shown in Fig. 1 as computed at the optimum geometries of the neutral and of the doubly positively charged oligomer respectively. The computed energies are referred to that of the neutral from at its optimized geometry. Comparison between the 3PDAs and 15PDAs results clearly shows the strong energetic (i.e. structural) effects due to the bipolaron “confinement”. The geometry relaxation stabilizes the doubly charged oligomer by 1.07 eV for 3PDAs, but by only 0.2 eV for 15PDAs, and with respect to the neutral form these structural modifications imply an energy increase of 4.18 eV and only 0.26 eV for 3PDAs and 15PDAs respectively. These facts are a consequence of the constrained localization of charge in the shorter oligomer and of the possibility of having a polaron pair (with a consequent increased charge delocalization) in the larger one. In order to discuss the thermodynamic stability of a bipolaron/polaron pair versus

Evolution of the Bipolaron Structure in Oligo-diacetylene Films 15PDAs 3PDAs { } = Optimized (AM1) geometry [ ] = Oligomer charge

[+]

[0]

[+]

{N}

{++}

{N}

135

1.75 eV [0]

[++]

[0]

{N}

{++}

{N}

-0.93 eV [+] {N}

[0] {++}

-1.19 eV

[+] {N}

[+]

[0]

[+]

{N}

{N}

{N}

-1.24 eV

(i)

(int)

[+]

[0]

[+]

{N}

{N}

{N}

(f)

Fig. 2. Schematic energy profile for the isodesmic reaction: 2 nPDAs + nPDAs[++] Φ 2 nPDAs[+,·] + nPDAs.

two radical cations, and the efficiency of the disproportionation reaction, consider the reaction path depicted in Fig. 2. The energy profile shows that the formation of two polarons on different chains is energetically favored with respect to that of a bipolaron/polaron pair by 56.1 (2.43) and 27.4 (1.19) Kcal·mol–1 (eV) for the 3PDAs and 15PDAs, respectively. The decreasing stabilization with increasing oligomer size is related to the intramolecular transformation of the bipolaron into a polaron pair [10,11] which implies a stabilization energy of 1.2 eV, see Fig. 1. The path proposed in Fig. 2 allows however for a further consideration which is of kinetic, rather than thermodynamic, nature. In fact, looking at the intermediate state there considered, it can be seen that for large oligomers (e.g., 15PDAs) the splitting of an intrachain polaron pair into single polarons on adjacent chains should occur with a very small activation energy. We should expect consequently that the formation of two polarons on different chains is a process whose reaction rate depends on the oligomer size, strongly increasing with increasing n. The above discussion does not explicitly take into account the relative orientation and the distance between the oligomer chains, as intermolecular interaction are usually neglected in theoretical studies. To obtain a more satisfactory model of the bipolaron behavior in bulk systems we have overcome this limitation. A detailed supramolecular study based on a semiempirical Hamiltonian is discussed elsewhere [12]. Here we only describe a simple electrostatic model of the physics of a doubly charged cluster, in which however almost the same conclusions can be reached. The model, aimed to describe the behavior of a film [13], is a cluster of three 15PDAs oligomers in a coplanar configuration as shown in Fig. 3. The central unit (II) has the optimized geometry of the doubly charged isolated chain, and the two neighbors that of the neutral isolated oligomer. The global charge of the cluster is assumed to be +2, so that for an infinite interchain distance the charge will be fully localized on the central unit.

136

M. Ottonelli, G. Musso and G. Dellepiane

1- δ+ δ

+

R'

R' R

R

δ+

R0/2 1- δ+ I

II

III

Fig. 3. Sketch of the electrostatic interaction between three oligodiacetylene chains. Unit II is the “doubly charged” one at the optimized geometry of the isolated doubly charged oligomer, and units I and III the “neutral” ones at the optimized geometry of the neutral.

To mimic the results of the full supramolecular calculation, we assume that at a fixed interchain distance R a process occurs in which a fraction 2δ of the positive excess charge is symmetrically transferred over the two “neutral” chains, and we analyze the first step of the process assuming point charges throughout. Under these conditions, it is shown (see Appendix) that the electrostatic interaction energy, for a fixed value of δ, can be repulsive ~or attractive depending on the fact that R is greater of less than a particular value R . On the other hand the amount δ of the excess charge which is transferred between the oligomer chains is itself a function of the strength of the interchain interaction (through the transfer integrals), that is of the interchain distance R. An unexpected result of our work [12] is that this R-dependence is of a somewhat complex nature, which is ~ revealed by the fact that δ goes through a maximum. This is because for R > R electrostatic effects favor the spread out of the excess charge on the neighbors chain, but at the same time increasing R tends to close the charge transfer channel by turning off the interaction. Closely (although not too mechanically) connected to the above behavior is the dependence of the interaction energy on R, which for frozen oligomer geometries can be seen to pass through a minimum, assuming the form of a Lennard-Jones-like curve. We remind that to the above electrostatic interaction energy a contribution should be added arising from a (partial) charge injection in the oligomer backbones due to their geometry relaxation inherent to the interaction. Acknowledgements. The research was supported by the Italian MIUR through the Fondo per gli Investimenti nella Ricerca di Base (FIRB 2001–2003) project.

References 1 2 3 4 5

G. Malliaras and R. H. Friend, Phys. Today 58, 53, 2005. G. Horovitz, Adv. Mater. 10, 365, 1998. A. S. Dhoot, G. M. Wang, D. Moses and A. J. Heeger, Phys. Rev. Lett. 96, 246403, 2006. K. E. Ziemelisetal, Phys. Rev. Lett. 66, 2231, 1991. J. L. Brédas, F. Wudl and A. J. Heeger, Solid State Commun. 63, 577, 1987.

Evolution of the Bipolaron Structure in Oligo-diacetylene Films 6 7 8 9 10 11 12 13

137

P. A. Lane, X. Wei and Z. V. Vardeny, Phys. Rev. Lett. 77, 1544, 1996. J. J. Apperloo, R. A. J. Janssen, Malenfant, P. R. L.; L. Groenendaal and J. M. Fréchet, J. Am. Chem. Soc. 122, 7042, 2000. C. Ehrendorfer and A. Karpfen, J. Phys. Chem. 98, 7492, 1994. H. Zuilhof, et al., in Supramolecular Photosensitive and Electroactive Materials, Edited by H. S. Nalwa, Cap. 4, 339, 2001. M. Ottonelli, I. Moggio, G. F. Musso, D. Comoretto, C. Cuniberti, and G. Dellepiane, Synth. Met. 124, 179, 2002. (a) G. Brocks, Synth. Met. 102, 914, 1999. (b) S. S. Zade and M. Bendikov J. Phys. Chem. B 110, 15839, 2006. M. Ottonelli, G. F. Musso and G. Dellepiane, J. Phys. Chem. A 112, 3991, 2008. M. Ottonelli, G. F. Musso, D. Comoretto and G. Dellepiane, Phys. Chem. Chem. Phys. 4, 2754, 2002.

Appendix Referring to Fig. 3 of the main text, the initial electrostatic energy (Ei) and that after a fraction 2δ of the excess charge is spread over the units of the cluster (Ef) are (in atomic units) Ei =

1 , R0

Ef =

δ2 (1 − δ ) 2 8δ (1 − δ ) + + R0 4R 2 + R 02 2R

(a.1)

and the electrostatic interaction energy is given by ΔE = −

δ (2 − δ ) ⎡ R0

⎤ δ2 ⎡ ⎤ 4 1 4 ⎢1 − ⎥+ ⎢ ⎥. − 2 2 ⎢⎣ 1 + (2R/R 0 ) ⎥⎦ R 0 ⎢⎣ (2R/R 0 ) 1 + (2R/R 0 ) ⎥⎦

(a.2)

If we neglect the second order terms (assuming δ << 1) equation a.2 simplifies in ΔE = −

2δ ⎤ 4 ⎥ 1− R0 ⎥ 1 + (2R/R 0 ) 2 ⎦

⎤ ⎥. ⎥⎦

(a.3)

In these conditions the electrostatic interaction energy becomes attractive when

ΔE < 0 ⇒ 1 −

4 1 + (2R/R 0 )

2

>0⇒R >

15 ~ R 0 = R ≅ 1.94R 0 . 2

(a.4)

In the opposite limit condition (δ ~ 1) the electrostatic interaction energy (from equation a.2) becomes attractive when ~ 1 ΔE < 0 ⇒ R > R = R 0 , 2

(a.5)

~ i.e. the value of R tends to decrease with increasing δ, which in turn varies with intermolecular distance (see the main text).

Molecular Orientation Dependence of the Ionization Energy of Pentacene in Thin Films Georg Heimel and Norbert Koch Humboldt-Universität zu Berlin, Institut für Physik, Newtonstraße 15, 12489 Berlin, Germany E-mail: [email protected] Abstract. An isolated individual molecule clearly has only one ionization energy. For ordered molecular assemblies, however, multiple values have been found depending on the orientation of the molecules relative to a supporting substrate. This intriguing observation is rationalized here for the prototypical molecule pentacene, in terms of intrinsic “surface dipoles” built into the molecules, which collectively give rise to the orientation dependence of the molecular ionization energy.

1.

Introduction

The work function (φ) is defined as the energy difference between the Fermi level (EF) and the vacuum level above a sample (Evac).It is known that φ of metals depends on the crystal face exposed to vacuum [1-3], e.g., φ spans a range of 0.5 eV for copper (100), (110), and (111) surfaces [1,2]. As EF is constant, this observation has been explained by crystal face dependent intrinsic “surface dipoles”. Differences in the geometric and, consequently, electronic structure cause a different amount of the electronic cloud to spill out of the bulk into the vacuum [3,4]. The resulting dipoles change Evac and thus impact φ. For molecular van der Waals crystals of non-polar molecules, surface dipoles and work function anisotropy have only recently been explored [5]. While variations of several tenths of an eV in the ionization energy (IE; the molecular analog to a metal’s φ) depending on the molecular orientation on a surface have been reported several times [6–8], a consistent picture was lacking. An explanation for the intriguing observation that one and the same molecule can have different – still well-defined – IEs in ordered thin films is that the electrostatic potential above a molecular crystal surface is determined by the orientation of the molecules and their intramolecular charge distribution. Conjugated organic molecules are of interest for applications in several optoelectronic devices, such as light emitting diodes, photovoltaic cells, or field effect transistors. As pentacene is the prototypical organic molecule for the use in organic thin film field effect transistors, in the present work we focus on demonstrating and explaining the orientation dependence of the IE of pentacene in thin films. Ultraviolet photoelectron spectroscopy (UPS) results for thin films with different molecular orientation are presented and compared to density functional theory (DFT) calculations on corresponding model molecular crystal surfaces. In essence,

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_23, © Springer-Verlag Berlin Heidelberg 2009

141

142

G. Heimel and N. Koch

we show that the ionization energy of pentacene varies by ca. 0.75 eV between flat lying and almost vertically standing orientations, which needs to be considered in the discussion of charge carrier injection barriers at interfaces of pentacene to electrodes in devices.

2.

Methodology

The UPS spectra of lying pentacene (PEN) on Au(111) [9] and standing PEN on a conducting polymer substrate [10] [poly(3,4-ethylenedioxythiophene)/poly (styrenesulfonate) – PEDT:PSS] have been reported before; details on sample preparation and measurements can be found in these earlier reports. In the DFT calculations the repeated slab approach was employed with a vacuum region separating two consecutive molecular layers by 20 Å. The PW91 exchange-correlation functional was used. For the valence-core interactions, the projector augmented-wave method [11,12] was employed permitting the low kinetic energy cutoff of 20 Ryd for the plane-wave expansion of the valence Kohn-Sham orbitals. Monkhorst-Pack [13] grids of 1 × 4 k-points (lying pentacene), and 4 × 4 (standing pentacene) were used for the integration of the 2D Brillouin zone. The atomic positions within the molecules were optimized until all remaining forces were < 0.01 eV/Å. All calculations were performed with the VASP code [14]. The following surface unit cell of PEN on Au(111) was used in the DFT calculations: a = 5.76 Å, b = 15.3 Å, γ = 79.1° [15]. Since also the actual thin film structures for the standing layers on PEDT:PSS are not known directly from experiment, a single layer of standing molecules was cut out of the PEN bulk structures for the DFT calculations; the lateral unit cell containing two molecules arranged in typical herringbone fashion was taken to be a = 6.266 Å, b = 7.742 Å, γ = 84.68° [16]. Due to intrinsic shortcomings of DFT, the calculated HOMO (highest occupied molecular orbital) energies usually underestimate IEs. As photo-hole screening is not included in standard DFT calculations, our calculated shifts in IE have to be regarded as shifts of the initial electronic states prior to removal of the photo-electron.

3.

Results and Discussion

Earlier studies have shown that the monolayer of PEN on Au(111) is lying flat on the surface [9,15], whereas an almost vertically standing molecular orientation prevails on the conducting polymer PEDT:PSS [10]. The corresponding UPS spectra are shown in Fig. 1 on a binding energy (BE) scale referenced to Evac. For both samples, the photoemission peak attributed to the PEN HOMO can be clearly seen at the low BE side of the spectra. For PEN on Au(111), additional photoemission intensity from the Au substrate is observed at BEs lower than the PEN HOMO (below 5.35 eV), exhibiting the characteristic metal Fermi-edge at 4.7 eV. Commonly, the IE of molecular materials is determined in UPS from the low-BE edge of the HOMO emission. Therefore, the IE of lying PEN (on Au) is 5.35 eV, whereas the IE of standing PEN (on PEDT:PSS) is only 4.60 eV. Consequently, IE of PEN layers with different molecular orientation is not constant, but differs by ΔIE = 0.75 in the present case.

Molecular Orientation Dependence of the Ionization Energy 143

intensity (arb. units)

standing PEN on PEDT:PSS

lying PEN on Au(111)

ΔIE= 0.75 eV

-8

-7

-6

-5

binding energy (eV), Evac= 0 eV

Fig. 1. UPS spectra for a lying PEN monolayer on Au(111) (bottom curve) and standing PEN on PEDT:PSS (top curve). The binding energy is given with respect to the vacuum level (Evac) set to zero.

This observation of orientation dependent IE values for PEN parallels that found for related rod-like molecules, such as sexithiophene and α,ω-dihexyl-sexithiophene [5], where ΔIE values of similar magnitude were reported. In order to understand the remarkable finding of PEN having a different IE depending on its orientation in an ordered structure, we performed DFT calculations on single layers of standing and lying PEN molecules (see Methodology section for details). The valence density-of-states (DOVS) for a lying (full line) and standing (dotted line) layer of PEN is shown in Fig. 2. In agreement with the experimental UPS spectra (Fig. 1), we find that all molecular levels are closer to Evac for the standing layer compared to the lying layer, i.e., the IE is lower for standing molecules. In addition to the good qualitative agreement between experiment and theory, also the quantitative agreement is very satisfactory; the calculations yield ΔIE = 0.7 eV (compared to the 0.75 eV found in the experiment). This result can be understood in terms of Evac being higher directly above the negatively charged π-electron system of PEN compared to the hydrogen-terminated ends of the molecule. To a first approximation, the potential outside the molecular layer is thus determined by the sum over the electron distributions of the individual molecules, giving rise to the pronounced orientation dependence of IE.

144

G. Heimel and N. Koch

-1

IEstanding

0 IElying

Potential Energy (eV)

Evac= 0

-2

-1 -2 -3

-3 HOMOstanding

-4 -5

HOMOlying

-4 -5 -6

-6 -7 lying

standing

DOVS

-7

Fig. 2. Results of the DFT calculations for the potential energy of the HOMO in a lying (left side) and a standing (center) monolayer of PEN. The right side shows the DOVS of the two layers with different orientation. Energy reference is the vacuum level (Evac).

4.

Conclusions

We have shown that the ionization energy of pentacene depends strongly on the orientation of molecules at the surface of ordered layers. While a lying layer of PEN had an IE of 5.35 eV, the IE value was only 4.6 eV for PEN standing layers. DFT calculations showed that the IE for a certain orientation of molecules within a layer is determined by an “intrinsic” molecular surface dipole, which results from the electron distribution of individual molecules within layers. The fact that IE values of ordered molecular layers and crystals depend on the surface molecular orientation needs to be taken into account when evaluating organic/electrode and organic/ organic interface energy levels for the optimization of organic electronic devices. Acknowledgements. The authors thank Steffen Duhm, Ingo Salzmann, Antje Vollmer, Jürgen P. Rabe, and Robert L. Johnson for experimental help and valuable discussions. NK acknowledges financial support by the Emmy Noether Program (DFG).

References 1 2

H. B. Michaelson, J. Appl. Phys. 48, 4729, 1977. H. L. Skriver and N. M. Rosengaard, Phys. Rev. B 46, 7157, 1992.

Molecular Orientation Dependence of the Ionization Energy 145 3 4 5 6 7 8 9 10 11 12 13 14 15 16

R. Smoluchowski, Phys. Rev. 60, 661, 1941. N. D. Lang and W. Kohn, Phys. Rev. B 3, 1215, 1971. S. Duhm, G. Heimel, I. Salzmann, H. Glowatzki, R. L. Johnson, A. Vollmer, J. P. Rabe and N. Koch, Nat. Mater. 7, 326, 2008. N. Koch, I. Salzmann, R. L. Johnson, J. Pflaum, R. Friedlein and J. P. Rabe, Org. Electron. 7, 537 2006. H. Fukagawa, H. Yamane, T. Kataoka, S. Kera, M. Nakamura, K. Kudo and N. Ueno, Phys. Rev. B 73, 245310, 2006. J. Ivanco, B. Winter, F. R. Netzer and M. G. Ramsey, Adv. Mater. 15, 1812, 2003. N. Koch, A. Vollmer, S. Duhm, Y. Sakamoto and T. Suzuki, Adv. Mater. 19, 112, 2007. N. Koch, A. Elschner, J. P. Rabe and R. L. Johnson, Adv. Mater. 17, 330, 2005. P. E. Blöchl, Phys. Rev. B 50, 17953, 1994. G. Kresse and D. Joubert, Phys. Rev. B 59, 1758, 1999. H. J. Monkhorst and J. D. Pack, Phys. Rev. B 13, 5188, 1976. G. Kresse and J. Furthmüller, Comp. Mater. Sci. 6, 15, 1996. C. B. France, P. G. Schroeder, J. C. Forsythe and B. A. Parkinson, Langmuir 19, 1274, 2003. C. C. Mattheus, A. B. Dros, J. Baas, A. Meetsma, J. L. de Boer and T. T. M. Palstra, Acta Cryst. C 57, 939, 2001.

Charge Transfer and Polarization Screening at Organic/Metal Interfaces: Single Crystalline Versus Polycrystalline Gold Heiko Peisert1, Daniel Kolacyak1, Andre Petershans1,2 and Thomas Chassé1 1

Institute of Physical and Theoretical Chemistry, University of Tübingen, Auf der Morgenstelle 8, 72076 Tübingen, Germany E-mail: [email protected] 2 Institute for Technical Chemistry, Water- and Geotechnology Division, Forschungszentrum Karlsruhe GmbH, D-76021 Karlsruhe, Abstract. We studied electronic polarization effects at organic/metal interfaces using combined photoemission spectroscopy (PES) and x-ray excited Auger electron spectroscopy (XAES) as a function of the organic layer thickness. As a model system for organic semiconductor/metal interfaces, magnesium phthalocyanine was evaporated onto gold. The influence of the morphology and orientation was studied by comparing different substrates: polycrystalline gold foil and single crystalline Au(100). Two different features in the Mg KLL spectra can be clearly separated for (sub-)monolayer coverages while only minor changes of the shape of Mg 1s are observed. Applying a dielectric continuum model, layer dependent screening contributions are estimated for metal-dielectric interfaces. The major screening mechanism cannot be described sufficiently by polarization screening due to mirror charges, significant contributions by charge transfer screening have to be considered.

1.

Introduction

In organic semiconductors the electronic polarization of the dielectric medium by charge carriers is a fundamental property of their electronic structure since the polarization energy is usually greater than transfer integrals or thermal energy. The polarization affects therefore the charge carrier transport in organic materials significantly. In particular, at surfaces and interfaces the electronic polarization is different compared to the bulk organic material [1], which may affect the charge carrier injection [2,3]. In this work we study electronic polarization effects at organic/metal interfaces, utilizing the screening mechanism of holes in final states of photoemission and Auger spectroscopy. As a model system we choose magnesium phthalocyanine (MgPc) on gold, a system where no chemical interaction occurs. Experiments were done on Au(100) and compared to recently published data on polycrystalline gold foil [4], in order to gather information about the influence of the molecular ordering and orientation on the observed effects.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_24, © Springer-Verlag Berlin Heidelberg 2009

147

148

H. Peisert et al.

2.

Experimental Methods

The interfaces were studied using core level X-ray photoemission spectroscopy (XPS), x-ray excited Auger electron spectroscopy (XAES) and valence band ultraviolet photoemission spectroscopy (UPS). The measurements were performed using a multi-chamber UHV-system (base pressure 2·10–10 mbar), equipped with a Phoibos 100 hemispherical analyzer (SPECS), a monochromatic Al Kα source and a He discharge lamp. Organic films were evaporated in a step-wise manner onto Au(100). The quality of the surface preparation was checked by LEED (low energy electron diffraction). The pressure during the evaporation of the PCs was less than 5 × 10–8 mbar, the evaporation rate was about 0.01 nm/s. The film thickness was controlled by a quartz microbalance calibrated using the attenuation of XPS intensities during the initial deposition steps.

3.

Results and Discussion

In photoemission spectroscopy, the removal of an electron affects the whole electron system, the remaining electrons of the environment screen the photohole. Due to the different final states in XPS (one hole) and XAES (two holes) such final state (FS) effects influence in particular shifts in Auger spectra and thus the comparison of energetic shifts in XPS and XAES enables the estimation of electronic relaxation energy ΔRD (screening ability) via the Auger parameter [4]. Presuming similar intra-molecular screening and excluding extra-molecular charge transfer within the time scale of the photoemission, ΔRD can be correlated with the change of the polarization energy induced by the redistribution of environmental charges. In Fig. 1 we compare Mg 1s photoemission core level spectra and Mg 1s

Mg KL2,3L2,3

thickness (nm)

1S

1D

thickness (nm) 10.8

2.4

1.1

Intensity (arb. units)

Intensity (arb. units)

10.8

2.4

1.1

0.3

0.3 1.8 eV Au

0.8 eV

1306

1304

1302

Binding energy (eV)

1300

Au

312 310 308 306 304 302 300 Binding energy (eV)

Fig. 1. MgPc/Au(100): Mg 1s photoemission spectra (left) and Mg KLL Auger spectra (right) of incrementally deposited MgPc on single crystalline Au. The well resolved additional feature at the earliest stages of deposition appears only in the Auger spectra.

Charge Transfer and Polarization Screening at Organic/Metal Interfaces

149

Mg KLL(1D) Auger spectra of incrementally deposited films of MgPc on single crystalline Au(100). In both cases a strong shift of spectral features to higher binding energies (EBs) with increasing film thickness can be observed as expected for ultrathin films <2nm [5]. However, for ultrathin films in the monolayer range two different features in the Mg KLL spectra can be clearly separated while only minor changes of the shape of Mg 1s are observed. This different behavior suggests an attribution predominantly to screening, since the contribution of ΔRD to energetic shifts of Auger spectra is expected to be threefold higher compared to XP spectra [4]. Thus, a splitting of photoemission signals may not be energetically resolved, it could be only visible as a broadening of the spectra. From the analysis of the Auger peak ratio at 1.1 nm (Fig. 1) we obtain that the difference of the screening is mainly related to the first molecular layer at the interface (assuming lying molecules [6] and a molecule-molecule distance of

Modified Auger parameter α' (eV)

MgPc/Au foil 2487.5

interface signal 2487.0

~1.8 eV

2486.5

bulk signal

2485.5 2485.0 0

2

4

6

8

10

12

14

Layer thickness (nm)

Modified Auger parameter α' (eV)

MgPc/Au(100) 2487.5

interface signal

2487.0

~ 1.8 eV

2486.5 2485.5

bulk signal 2485.0 0

2

4

6

8

10

12

14

Layer thickness (nm)

Fig. 2. Comparison of modified Auger parameters for MgPc on Au foil and on Au(100). We distinguish between interface and bulk signal, the average difference Δα’ for the two components (1.8 eV) is similar in both cases.

150

H. Peisert et al.

about 0.34 nm). These findings agree well with the system MgPc/Au foil [4]. Thus, a change of the orientation, as observed for phthalocyanines on gold foil at a thickness of about 2–3 monolayers [7], is not responsible for the observed changes in the screening of the XPS/XAES final states. In Fig. 2 we compare modified Auger parameters α’ (sum of the kinetic energy of the Auger and the BE of the XPS peak) for MgPc on gold foil (top) and on Au(100) (bottom). In both cases we distinguish clearly between interface and bulk signals, the distance Δα’ is about 1.8 eV. We note that a possible separation of the photoemission signals into bulk and interface is not considered (see above, Fig. 1). From the peak fitting analysis of Mg 1s and of the HOMO (UPS) we estimate an upper limit for a possible splitting of 0.3 eV (see also [4] for Au foil), i.e. in this case Δα’ may be lowered by 0.3 eV. The estimation of the relaxation energy for molecules directly at the metal surface and molecules in thicker films via the Auger parameter (Δα’ = 2ΔRD) yield to ΔRD ~ 0.9 eV for MgPc on Au foil and on Au(100). On the other hand, layer dependent screening contributions can be estimated for metal-dielectric interfaces applying a dielectric continuum model according to ΔEB(d ) − ΔEB(∞ ) = −e 2 / (16πε 0ε ⋅ d ) [4, 8], where d is the distance from the mirror plane, ΔEB(d) and ΔEB(∞) is referred to the distance d and the infinitely thick film, respectively. Here, we assume ε ~ 3 and 0.34 nm for the molecule-moleculedistance. The distance of the first layer to the mirror plane of the metal d1 could be different on a microscopic scale. We apply the van der Waals radius of carbon in organic compounds (analogously to [8]) d1 = 0.17 nm and for comparison a distinct larger value (0.23 nm). The results are summarized in Table 1: Table 1. Layer dependent screening contributions estimated by dielectric continuum model. For the distance of the first layer to the mirror plane of the metal 1.7 Δ (left) and 2.3 Δ are chosen.

1 layer 2nd layer 3rd layer st

d (Δ) 1.7 5.1 8.5

ΔRD (eV) 0.7 0.25 0.15

d (Δ) 2.3 5.7 9.1

ΔRD (eV) 0.52 0.21 0.13

Clearly visible from table 1 is, that the difference of ΔRD between the first and the second layer does not depend distinctly on d1. Resulting values are significantly lower (table 1, ~0.45 eV and 0.31 eV) as compared to the experimental observations. Moreover, in order to explain the experimental ΔRD, a distance of the first layer to the mirror plane is much smaller than the van der Waals radius has to be assumed, which is not reasonable. Thus the observed splitting in the Auger signals could be understood only partly by polarization screening, an additional contribution by charge transfer screening has to be considered. Since an electron transfer to the molecule is the most effective screening mechanism, large energetic shifts can be expected. Thus, the different screening behavior in particular for the first monolayer may be caused by an ultrafast charge transfer from the metal to the molecule, which affects in particular the Auger spectra. Femtosecond charge transfer dynamics were recently

Charge Transfer and Polarization Screening at Organic/Metal Interfaces

151

observed for other organic interfaces (e.g., organic materials on metals [9] or on MoS2 [10]) Thus, our experimental study may support recent studies where ultrafast charge transfer processes and/or induced interface states [11] are proposed. Acknowledgements. We are grateful to I. Biswas and W. Neu for valuable discussions and technical support. The work was supported by the German Research Council Ch 132/20-1.

References 1 2 3 4 5

E. V. Tsiper, Z. G. Soos, W. Gao, and A. Kahn, Chem. Phys. Lett. 360, 47, 2002. N. Koch, A. Vollmer, S. Duhm, Y. Sakamoto, T. Suzuki, Adv. Mater. 19, 112, 2007. X.-Y. Zhu, Surface Science Reports 56, 1, 2004. H. Peisert, A. Petershans, T. Chassé, J. Phys. Chem. C 112, 5703, 2008. H. Peisert, T. Schwieger, J. M. Auerhammer, M., Knupfer, M. S. Golden, and J. Fink, J. Appl. Phys. 91, 4872, 2002. 6 H. Peisert, T. Schwieger, J. M. Auerhammer, M. Knupfer, M. S. Golden, J. Fink, P. R. Bressler, and M. Mast, J. Appl. Phys. 90, 466, 2001. 7 I. Biswas, H. Peisert, M. Nagel, M. B. Casu, S. Schuppler, P. Nagel, E. Pellegrin, T. Chassé, J. Chem. Phys. 126, 174704/1–5, 2007. 8 T.-C. Chiang, G. Kaindl, G. Mandel, Phys. Rev. B 33, 695, 1986. 9 S. Neppl, U. Bauer, D. Menzel, P. Feulner, A. Shaporenko, M. Zharnikov, P. Kao, and D.L. Allara, Chem Phys. Lett. 447, 227 2007. 10 M. P. de Jong, R. Friedlein, S. L. Sorensen, G. Öhrwall, W. Osikowicz, C. Tengsted, S. K. M. Jönsson, M. Fahlman, W. R. Salaneck, Phys. Rev. B 72, 035448, 2005. 11 H. Vázquez, Y. J. Dappe, J. Ortega, and F. Flores, J. Chem. Phys. 126, 144703, 2007.

Sensing Infrared Light With an Organic/Inorganic Hetero-Junction Gebhard J. Matt1, Thomas Fromherz1 , Guillaume Goncalves2, Christoph Lungenschmied3, Dieter Meissner 4 and Serdar N. Sariciftci4 1

Institute for Semiconductor and Solid State Physics, Johannes Kepler University Linz, Altenbergerstrasse 69, 4040, Austria E-mail: [email protected] 2 Ecole Nationale Superieure de Chimie et de Physique de Bordeaux (ENSCPB) E-mail: [email protected] 3 Konarka Technologies, Johannes Kepler University Linz, Altenbergerstrasse 69, 4040, Austria E-mail: [email protected] 4 Linz Institute for Organic Solar Cells, Johannes Kepler University Linz, Altenbergerstrasse 69, 4040, Austria E-mail: [email protected] Abstract. Organic and inorganic semiconductors are diverse in many of their physical properties but the combination of these can feature unexpected as well as unique physical properties. Here we report on an organic/inorganic hetero-junction which can readily be utilized for sensing infrared light. The used materials are highly boron doped crystalline silicon (100) and a soluble buckminster fullerene derivative. Despite the missing light absorption of silicon as well as the of the fullerene derivative for a wavelengths beyond 1.3 μm and 720 nm respectively, a hetero-junction of these materials absorbs in the infrared and generates a primary photo-current in the wavelength range from 1.3 to 3 μm. The simple preparation of the hetero-junction by a solution process of the fullerene derivative on top of the silicon wafer substrate, is technologically attractive in an economically viable way.

1.

Introduction

Here we report on a novel light sensing device based on a silicon/fullerene hetero– junction that allows the realization of optoelectronic devices for the NIR which are fully compatible with CMOS technology. In essence, the inherent disadvantage of silicon for optoelectronic infrared applications is its transparency beyond a wavelength of 1.1 μm. To overcome this disadvantage, several technologies such as the hetero–epitaxial growth of (polycrystalline-) germanium on silicon or the usage of in near infrared photo–conductive and soluble nano–particles [1] have been developed. In the latter case, the facile solution processing is of particular interest. Potential soluble semiconductors are organic conjugated polymers and fullerene derivatives. The strengths of this material – class are the solubility, excellent optical properties and the per se tunable chemically structure. In this work, the soluble C60 derivative methano-fullerene [6,6] phenylC61 butyric acid methyl ester (PCBM) has been chosen (see left graph in Fig. 1 where the chemical structure is depicted).

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_25, © Springer-Verlag Berlin Heidelberg 2009

153

154

G.J. Matt et al.

In contrast to C60, PCBM is soluble up to 5 weight percent in common organic solvents. In this paper it is shown that a p-Si/PCBM hetero–junction features a photo– voltaic effect in the infrared regime between photon energies from 0.4 to 1.1eV (3–1.3 μm) [2].

Fig. 1. Left graph shows the chemical structure of the fullerene derivative – PCBM. Right graph shows a schematic cross section of the Al/p–Si/PCBM/Al hetero–junction with an sample photograph [2].

2.

Experimental Methods

The investigated samples have a layered structure (right graph in Fig. 1) and the hybrid hetero-junction is manufactured by spin-coating a PCBM solution (3 weight% in chlorobenzene) on top of a p-Si substrate. The resulting PCBM film thickness is 140 nm. Subsequently a 100 nm thick aluminum contact is evaporated under dynamic vacuum (10–6 mbar) on top of the PCBM thin-film as well as directly to the p-Si substrate. For the electrical characterizations a Keithley 236 SMU has been used. The photo-current spectra were measured with a Bruker 113v Fourier transform spectrometer. The short circuit photo-current generated in the sample was amplified by a Stanford Research Systems SR570 low-noise current amplifier and fed back into the spectrometer electronics via the input for external detectors.

3.

Results

In Fig. 2 the current-density versus voltage (IV) characteristics are presented at room temperature and at 80 K. At 295 K the IV-characteristics exhibit a current rectification ratio of 8.105 for a bias variation from –1 to +1 V. Upon cooling, the reverse dark current-density at -2 V bias decreases from 5.10–7 A/cm2 at 295 K to the sub nA/cm2 region at 80 K. From an Arrhenius plot of the reverse dark current-density

Sensing Infrared Light With an Organic/Inorganic Hetero-Junction

155

Fig.2. Current-density versus voltage (IV) characteristics of an Al/p-Si/PCBM/Al heterojunction at 295 K (top graph) and at 80 K (bottom graph). The inset shows an Arrhenius plot of the reverse current-density at –2V bias. In the bottom graph the IV-characteristics at 80 K under broadband IR illumination (red triangles) is shown [2].

at –2 V (see inset in the top graph of Fig. 2), an activation energy of Ea = 0.3 eV is found. In dark and at 80 K, the IV-characteristics features two distinct current minima, at –0.5 V and at +0.5 V bias voltage for opposite sweep directions. This hysteresis behavior is attributed to trapped charges in the PCBM thin-film. At a sample temperature of 80 K and under broadband NIR illumination from a tungsten lamp spectrally restricted by a Si filter, an IV-characteristics typically for a photo–voltaic device is observed (see bottom graph in Fig. 2). Under the same experimental conditions as for the IV measurements, the NIR photo-current at various temperatures is spectrally resolved using a Fourier–transform spectrometer. Above 220 K sample temperature, a photo-current solely around 1.1eV is observed

156

G.J. Matt et al.

(red graph in Fig. 3). Reducing the sample temperature, the photo–current at 1.1 eV is decreasing, and a spectral monotonically increasing photo–current in the energy range from 0.4eV to the cut-off of the at room temperature maintained Si filter is measured. Upon cooling the signal to noise ratio increases and saturates at temperatures below 150 K. The onset and the spectral shape of the photo–current between 0.4 and 1.1eV is only weakly temperature dependent. The interaction of fullerenes with semiconductors and metals is strong and complex. In particular, fullerenes on bare silicon are very likely to be chemisorbed which entails a charge transfer and accounts for an additional bonding beyond van der Waals bonding [3,4]. Low work-function metals as Al or Ca, form an ohmic electron injecting contact to fullerene thin-films [5]. As Al is also ohmic to p-Si too, it is concluded, that the observed rectification if the IV-characteristics is due to the p-Si/fullerene interface.

Fig. 3. Spectrally resolved photo-current of an Al/p-Si/PCBM/Al hetero-junction in the temperature range of 220 to 80 K. The dotted line in the top graph shows the transmission of the at room temperature kept silicon filter [2].

Under an applied positive (forward) bias voltage, electrons are efficiently injected from the ohmic Al top–contact into the PCBM layer and recombine with holes in the p-Si. When biasing the p-Si/PCBM diode in reverse direction, holes are extracted from the p-Si valence band (VB) into the Al back-contact. But this holecurrent is limited by the hole blocking properties of the PCBM thin-film and the Al top contact receptively. Out of the temperature dependence of the reverse dark current between 305 K and 205 K, an activation energy Ea of 0.3 eV is determined from the Arrhenius plot, shown in the inset of Fig. 2. This result indicates and

Sensing Infrared Light With an Organic/Inorganic Hetero-Junction

157

manifests a strong interaction of the PCBM–LUMO with the Si VB states across the interface. Ea is interpreted as the height of a band-offset (barrier) between the Si VB and the LUMO of the PCBM thin-film. The strong PC at 1.1 eV is assigned to the excitation of an electron from the Si VB to the CB and its subsequent injection into the PCBM. Due to the transparency in the spectral range below 1.17 eV of the silicon and PCBM respectively, the photo-current response between 0.4eV and 1.1eV cannot be trivially assigned to a direct absorption in either of the materials. Instead, it is ascribed to an (in real space indirect) optical transition from the VB of p-Si to the LUMO of the PCBM thin-film. The onset of the IR photo-current at 0.4 eV is consistent with by the Arrhenius plot estimated barrier height Ea of 0.3 eV.

4. Conclusions In summary, it has been shown that by infrared radiation charge-carriers can be directly excited across the interface of a silicon/fullerene hybrid hetero-junction. Besides its scientific relevance, the simple fabrication process as well as its compatibility with the well established silicon technology makes the presented hybrid approach a promising candidate for widespread applications. Acknowledgments. This work was supported by the Austrian Science Funds (FWF) and the Austria Wirtschaftsservice.

References 1. 2. 3. 4. 5.

G. Konstantatos, et al., Nature 442, 180 (2006). G. J. Matt, T. Fromherz, G. Goncalves, C. Lungenschmied, D. Meissner, N. S. Sariciftci and G. Bauer, submitted (2008) S. Suto, K. Sakamoto and T. Wakita, Physical Review B 56, 7439 (1997). T. R. Ohno, et al., Physical Rewiew B 44, 13747 (1991). C. J. Brabec, et al., Advanced Functional Materials 11, 374 (2001).

Ultrafast Confocal Microscope for Functional Imaging of Organic Thin Films Dario Polli, Michele Celebrano, Jenny Clark, Giulia Grancini, Tersilla Virgili, Guglielmo Lanzani and Giulio Cerullo Dipartimento di Fisica, Politecnico di Milano, Piazza L. da Vinci 32, 20133 Milano, Italy E-mail: [email protected] Abstract. We present a novel instrument combining femtosecond pump-probe spectroscopy with broadband detection and confocal microscopy. The system has 200-fs temporal resolution and 300-nm spatial resolution. We apply the instrument to map excited state dynamics in thin films of polyfluorene-polymethylethacrylate blends.

1.

Introduction

Ultrafast optical spectroscopy provides a wealth of information on the photophysics of organic semiconductors [1]. The photoinduced excited singlet states can undergo a variety of processes, such as mono- and bimolecular decay, internal conversion, energy and charge transfer or intersystem crossing to a triplet state. Understanding the dynamics and efficiency of such processes is crucial, both from a fundamental point of view and for optimizing the efficiency of organic optoelectronic devices, which are based on these effects. A peculiar feature of organic semiconductors, as compared to inorganic ones, is structural inhomogeneity; these materials, in the solid state, are generally amorphous or polycrystalline, with local order only achieved in mesoscopic domains with size ranging from a few tens to a few hundreds of nanometers [2]. In addition, many devices use blends of different molecules, which undergo phase separation into domains of varying size and shape [3]. The size and type of the mesoscopic structures has a decisive influence on excited state dynamics, which in turn determines fluorescence quantum yield, charge carrier mobility and generation efficiency. Femtosecond pump-probe spectroscopy experiments are usually performed over relatively large sample areas, with diameter of the order of several tens of microns, thus obtaining a macroscopic information that is averaged over many mesoscopic domains. An instrument combining the temporal resolution of pumpprobe spectroscopy with the spatial resolution of confocal microscopy would allow to study excited state dynamics of individual mesoscopic domains. The derived information would enable to establish a link between the observed dynamical optical properties and the environment in which they occur [4], providing an important feedback to material scientists who strive to control the morphology and the supramolecular organization of organic thin films in order to optimize the device performance. In this paper we present a novel instrument which combines femtosecond pump-probe spectroscopy with broadband detection and confocal microscopy. Our system enables to simultaneously achieve 200-fs temporal resolution and 300-nm H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_26, © Springer-Verlag Berlin Heidelberg 2009

161

162 D. Polli et al.

spatial resolution and thus to study excited state dynamics in organic semiconductors with a spatial resolution pertinent to their morphology. The measurements provide information not available with other microscopy techniques allowing a completely new insight into the mesoscopic structure of organic materials. After a brief description of the experimental setup, we present the first results on mapping excited state dynamics in a conjugated polymer blend. It is driven by 10-μJ, 150-fs pulses at f = 1 kHz repetition rate and 780 nm wavelength produced by an amplified Ti:sapphire laser. A fraction of the pulse energy is frequency doubled to generate the pump pulses at 390 nm; the remaining part is focused in a sapphire plate to produce a single filament white light continuum, used as a probe [5]. The pump beam is modulated by a chopper at f/2 and single probe colors are selected by 10-nm bandwidth interference filters in the 450−1000 nm wavelength range. Pump and probe pulses, synchronized by a computer-controlled delay line, are collinearly recombined by a dichroic beam splitter (BS2 in the figure) and focused on the sample by a microscope objective with 100 × magnification and 0.75 numerical aperture. The reflected probe, transmitted by BS3 and spectrally filtered to reject the pump light, is focused on the 50-μm core of an optical fiber, serving as the pinhole of the confocal microscope, and detected by a photomultiplier and two lock-in amplifiers. By raster scanning the sample, mounted on a piezotranslator, and referencing the first lock-in at f, one can acquire reflectance (R) maps at the probe wavelength, as in an ordinary confocal microscope. At the same time, by referencing the second lock-in at the pump modulation frequency f/2, one can measure the differential reflectance (ΔR/R) signal, defined as:

2.

Experimental Methods

The experimental setup of the ultrafast confocal microscope is shown in Figure 1.

Fig. 1. Experimental setup of the ultrafast confocal microscope. BS1 and BS3 are 50% beam splitters, while BS2 is a dichroic beam splitter. SHG: second harmonic generation crystal; OBJ: microscope objective; PZT: piezo translator; PMT: photomultiplier; IF: interference filter.

Ultrafast Confocal Microscope for Functional Imaging of Organic Thin Films ΔR R

=

R

pump on

R

− R pump off

163

.

pump off

One can either acquire ΔR/R maps for a given pump-probe delay τ or ΔR/R dynamics for a given spot on the sample. The instrument has a temporal resolution (given by pump-probe cross-correlation) of 200 fs and a spatial resolution (given by the point spread function of the microscope) of 300−400 nm, depending on probe wavelength. The minimum detectable signal is ΔR/R ≈10–4. Note that, due to the chromatic aberrations of the objective, pump and probe beams have different focal planes. In our experiment we optimize the focusing of the probe on the sample; the pump will be out of focus and will excite a larger portion of the sample than that actually probed (global pump), thus providing a spatially uniform excitation of the probed area.

Fig. 2. (a) absorption and photoluminescence spectra of a PFO thin film, (b) differential transmission spectrum for 390-nm pump and pump-probe delay τ = 1 ps.

3.

Results and Discussion

We have applied the ultrafast confocal microscope to map excited state dynamics in thin films of poly(9,9-dioctylfluorene) (PFO, see chemical structure in figure 2(a)), blended with polymethylmethacrylate (PMMA, 10% wt. PFO in PMMA). PFO is a blue-emitting polymer, with an absorption maximum at 385 nm (see Fig. 2(a)), while PMMA is transparent at our pump wavelength and it does not interact with PFO [6] so that it is optically inert. Figure 2(b) shows the macroscopic ΔT/T spectrum of PFO measured at τ = 1 ps; at 570 nm probe wavelength we observe a photo-induced absorption (PA) due to photo-generated polarons [7].

164 D. Polli et al.

Films with few-μm thickness were drop cast on a reflective metal substrate, so as to turn the ΔR/R into a double-pass differential transmission (ΔT/T). The presence of the metal surface does not affect the optical properties of PFO [7]. As is commonly the case in polymer blends, the film clearly shows a phase separation between PFO and the host matrix, with a large distribution of domain sizes ranging from sub-micron to hundreds of microns. Due to its rich spatial structure and the optical inertness of PMMA, this film is ideal to assess the performances of our ultrafast confocal microscope.

Fig. 3. (a) reflectance map of a PFO/PMMA blend at 570 nm wavelength; (b) ΔR/R map of the same sample area at 570 nm probe wavelength and τ = 200 fs; (c) same as (b) but at τ < 0.

Figure 3(a) shows the reflectance map of a portion of the sample at 570 nm probe wavelength. The contrast is rather low because neither PFO nor PMMA absorb at this wavelength and can be attributed to local variations in scattering. Figure 3(b) shows the ΔR/R map at τ = 200 fs: we observe a strong difference between PMMA (which displays zero signal) and PFO (which displays a PA signal). Note that the contrast of the ΔR/R image is much higher than that of a standard confocal reflectance map and the spatial resolution is also higher. The ΔR/R signal allows to observe clearly the distribution of PFO in the blend. Figure 3(c) shows the ΔR/R map at τ < 0; in this case, the signal is zero everywhere since we are probing an unpumped sample. Measuring ΔR/R dynamics at various positions on the sample, different decay times are found, dependent on the local environment of PFO [8].

Ultrafast Confocal Microscope for Functional Imaging of Organic Thin Films

4.

165

Conclusions

In this work we have introduced a novel instrument for investigating the photophysical properties of thin solid films: an ultrafast confocal microscope. The system enables to measure transient absorption dynamics with a combination of high temporal (200 fs) and spatial (300 nm) resolution. Our technique allows to probe two additional dimensions with respect to standard confocal microscopy: in fact we can acquire ΔR/R maps at different pump-probe delays, thus highlighting differences in the lifetimes of the photexcitations. In addition by changing the probe wavelength we can map processes such as charge generation, energy transfer, intersystem crossing etc. The results will allow the representation of excited state dynamics in organic thin films by multi-dimensional maps in real and phase space.

References 1 2 3 4 5 6 7 8

G. Lanzani, G. Cerullo, D. Polli, A. Gambetta, M. Zavelani-Rossi, C. Gadermaier, Phys. Stat. Sol. 201, 1116 (2004). B.J. Schwartz, Annu. Rev. Phys. Chem. 54, 141 (2003). J.J. M.Halls, C.A. Walsh, N.C. Greenham, E.A. Marseglia, R.H. Friend, S.C. Moratti, A.B. Holmes, Nature 376, 498 (1995). A. Cadby, R. Dean, A.M. Fox, R.A.L. Jones, D.G. Lidzey, Nano. Lett. 5, 2232 (2005). G. Cerullo, S. Stagira, M. Nisoli, S. De Silvestri, G. Lanzani, G. Kanzelbinder, W. Graupner, G. Leising, Phys. Rev. B 57, 12806 (1998). J. Chappell, D.G. Lidzey, J. Microsc. 209, 188 (2003). T. Virgili, G. Cerullo, C. Gadermaier, L. Lüer, G. Lanzani, D.D.C. Bradley, Phys. Rev. Lett. 90, 247402 (2003). D. Polli, J. Clark, M. Celebrano, G. Grancini, T. Virgili, G. Lanzani, G. Cerullo, manuscript in preparation.

Growth and Desorption Kinetics of Sexiphenyl Needles: an in-situ AFM/PEEM Study Alexander J. Fleming, Svetlozar Surnev, Falko P. Netzer, and Michael G. Ramsey Surface and Interface Science, Dept. of Physics, Karl-Franzens University, Graz A-8010, Austria. E-mail: [email protected] Abstract. The controlled in-situ deposition of sexiphenyl (6P) on the (2x1) oxygen reconstruction of Cu (110) is shown to give rise to the ordered growth of large anisotropic needle-like structures on the surface. Photoemission electron microscopy (PEEM) and atomic force microscopy (AFM) results are presented for the growth of 6P (20-3) crystalline needles for a range of substrate temperatures. In addition, desorption and other interesting phenomena are discussed.

1.

Introduction, Results and Discussion

Anisotropic properties of surfaces and molecules are of particular interest for the development of low dimensional molecular nanostructures. In this regard the combination of PEEM and AFM enables the growth of 6P needles to be followed over a large scale range from nanometers to several tens of micrometers. 6P on Cu (110) (2 × 1) O is a well characterized system [1] both in terms of its electronic structure and crystalline structure and is found to be highly reproducible under UHV growth conditions. The PEEM image of 6P needles on Cu (2 × 1) O surface, given in Fig. 1a), shows the needles’ long axis (fastest growth direction) is perpendicular to the O rows. From X-ray diffraction studies it is known that 6P molecules in the (20-3) molecular crystal align parallel to the O rows. This indicates that the molecular crystal needles and molecular orientation are equivalent to those on TiO2 (110) and can thus be attributed to sticking anisotropy [2]. AFM images in Fig. 1 b) and c) clearly show that the needles are not single crystals but are rather comprised of several (20-3) crystalline grains with distinct inter-grain boundaries. As can be seen in Fig. 1 c) and d), the effect on the needle growth direction by steps and step bunches appears to be negligible. It is yet to be determined what exact role the steps and step direction of the Cu (2 × 1) O surface have in producing inter-grain boundaries in the needles. In any case, the anisotropy of the O rows and the 6P wetting layer guarantee the formation of highly ordered uni-axially oriented needles. In Fig. 2a) the evolution of the total photoelectron intensity, monitored during deposition of 6P, is given. The sharp rise in intensity up to one monolayer (ML) is due to a lowering of the workfunction below the energy of photons from the Hg lamp (4.9 eV). The increase in intensity is not monotonic but instead appears to progress at a different rate beyond 0.2 ML. This is attributed to a change in packing of the 6P wetting layer such as a change from flat-lying to tilted orientation. A further change is observed around 1.2 ML during the formation of a H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_27, © Springer-Verlag Berlin Heidelberg 2009

167

168

A.J. Fleming et al.

meta-stable second layer that leads to a decrease in the photoelectron intensity due to attenuation by second layer molecules. Beyond this second layer the intensity increases and dark needles are observed to start growing. This strongly indicates that a spontaneous depletion of the second layer forms the needles. All subsequent molecules deposited contribute to the needle growth. As a further test, the molecular beam shutter was opened and closed several times with an interval time of roughly 150s. The resultant changes in photoelectron intensity are shown in the inset of Fig. 2b). Repeated at several temperatures, the rise in intensity due to second layer depletion is plotted with respect to a)

c)

b)

d) oxygen rows

Normalised Intensity (a.u.)

b) φ ~ 0.2 eV

Intensity (a.u.)

a)

0

1

2

3

Normalised Time [monolayers]

4

1.0 0.8

Intensity [a.u.]

Fig. 1. a) PEEM image of 6P needles (scale bar 10 µm) illustrating the well ordered molecular structures. Note that the needles are dark and the wetting layer between the needles is bright. b) AFM frequency shift (Df) image reveals the inter-grain boundaries within the needle. The height of this needle is 50 nm. Needle growth over steps and step bunches (steps indicated by an arrow) in c) AFM Df image and d) PEEM image (10 µm wide).

6

7

8 9 10 Time [100 s]

11

0.6

RT 100°C 115°C 135°C 150°C

0.4 0.2 0.0

0

2

4

1/2

(Time)

1/2

6

8

[s ]

Fig. 2. a) Photoelectron intensity vs normalized time in ML, intensity scale bar indicates corresponding change in workfunction. Insets: PEEM images (20 µm) of surface taken at key stages of growth. b) Photoelectron intensity versus square-root-time at various temperatures. Inset: photoelectron intensity vs time at 115°C with molecular beam shutter periodically opened and shut.

Growth and Desorption Kinetics of Sexiphenyl Needles

169

square-root-time in Fig. 2b). The linearity of the RT measurement in the plot confirms that at this temperature the molecular diffusion is 1-D i.e. motion is confined along the O rows. At higher temperatures the increase is supra-linear, suggesting pseudo 2-D diffusion, most likely due to thermal activation of molecular hopping between O rows. The nucleation density, plotted as an Arrhenius plot in Fig. 3a), indicates that there is a significant energy barrier of 857 meV for the process of nucleation (for PTCDA on Ag(111) it is 740 ± 200 meV [3]). As shown in Fig. 3b), the binding energy of 6P molecules is also readily accessible from the slope of the Arrhenius plot of the rate of change of the needle length versus temperature during thermal desorption. The binding energy per 6P molecule is measured to be 2.1 ± 0.1 eV.

Fig. 3. a) Arrhenius plot of nucleation density. Inset: PEEM images (10 µm) of the growth at various temperatures. b) Arrhenius plot of 6P needle length contraction during thermal desorption to determine molecular binding energy. Inset: Snap-shot PEEM images of 6P desorption demonstrating the length contraction of the needle.

2.

Conclusions

It has been shown that PEEM is a versatile instrument that can be used to determine the thermodynamic and kinetic properties of molecular systems. The preliminary results presented here are presently being analysed to yield statistically significant values for the molecular binding energy as well as for the molecular surface diffusion constant and other kinetic parameters of interest. Acknowledgements: Supported by the Austrian Science Foundation (FWF).

References 1 2 3

G. Koller, S. Berkebile, M. Oehzelt, P. Puschnig, C. Ambrosch-Draxl, F. P. Netzer and M. G. Ramsey, Science, Vol. 317, 351 (2007) S. Berkebile, G. Koller, G. Hwalacek, C. Teichert, F. P. Netzer and M. G. Ramsey, Surface Science Letters, Vol. 600, 313 (2006) H. Marchetto, U. Groh, Th. Schmidt, R. Fink, H.-J. Freund and E. Umbach, Chem. Phys, Vol. 325, 178 (2006)

Temperature Dependence of the Charge Transport in a C60 based Organic Field Effect Transistor Mujeeb Ullah1, Th. B. Singh2, G. J. Matt2, C. Simbruner1, G. Hernandz-Sosa1, S. N. Sariciftci2, H. Sitter1 1 2

Institute of Semiconductor and Solid State Physics, JKU Linz, Austria. Linz Institute of Organic Solar Cells, JKU Linz, Austria.

Abstract. We studied the temperature dependence of the electron transport in C60 based Organic Field Effect Transistors (OFETs). On the spin-coated bottom gate dielectric, the semi-conducting C60 thin-film has been grown by standard evaporation technique. Device parameters as the threshold voltage, the field effect mobility and the activation energy of the electron transport were determined in the temperature range from 300 K to 77 K. The field effect mobility obeys the Meyer-Neldel Rule (MNR), which is an empirical relation between activation energy and the mobility prefactor.

1.

Introduction

The charge transport mechanism in organic field effect transistors (OFET) has been subject of research for some years now [1–11]. Although the temperature (T) and gate voltage (VG) dependence of the charge carrier mobility (μFE) of OFETs has been reported [1–12], the observed thermally activated behaviour of charge transport mechanisms in these organic devices is still not fully understood. It is even difficult to obtain an accurate picture of the nature of the transport due to large variations in the experimental data on even nominally the same samples [1–11]. Many organic materials have been reported to follow the Meyer-Neldel Rule (MNR) which is a phenomenological model to describe the observed temperature dependence of the mobility. The thermally activated behavior of the field effect mobility can be described as [3, 4, 7, 8, 10, 13–15] ⎛ −EA ⎞ μ = μ o exp ⎜ (1) ⎟ ⎝ k BT ⎠ where EA is the variable activation energy, T is the absolute temperature, kB is Boltzmann constant and µo is a prefactor, which can be found empirically. This prefactor also depends on EA. Equation (2) which gives the relation between the mobility prefactor (µo) and EA is called the MNR.

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_28, © Springer-Verlag Berlin Heidelberg 2009

173

174

M. Ullah et al.

⎛

EA ⎞ ⎟ ⎝ k B TM N ⎠

(2)

⎛ 1 1 ⎞⎤ − ⎟⎥ k B T M N ⎠ ⎥⎦ ⎝ k BT

(3)

μ o = μ oo exp ⎜ Inserting equation (2) into (1) gives:

⎡

μ = μ oo exp ⎢ − E A ⎜ ⎢⎣

which means, that at T = TMN the mobility becomes independent of EA and is called µMN. EA also depends on the applied VG. Consequently, if µ is plotted versus 1/T for different VG a common crossing point must be obtained at (µMN, 1/TMN). The MNR has been observed in a wide variety of physical, chemical and biological processes [4,7,8]. However, the microscopic origin of the MNR and therefore, the physical meaning of EMN, are still a topic of discussion [4, 7–9]. Among the n-type organic semiconductors, fullerenes with a symmetric structure and low ionization potential show a relatively high charge carrier mobility 0.6–6 cm2/ V–1s–1. [1, 5] Therefore we selected C60 based OFETs for our investigations.

2.

Experimental Methods

The device scheme of the OFET employed is shown in the inset of Fig. 1 (a). Details of device fabrication of the C60 OFETs are reported previously [1, 5]. The deposition of the C60 was done at RT using a Leybold Univex system at base pressure of 10–6 mbar. The completed devices were loaded in the Oxford cryostat under N atmosphere inside the glove box to avoid exposure to ambient conditions. The measurements were conducted in small temperature steps of 10K in the range from 300 K to 77 K. Every temperature step was kept 60 minutes so that the device got thermally stabilized. The output and transfer measurements were recorded by smu Agilent 2000.

3.

Results and Discussion

Figure 1. Shows the output characteristics of the device at 300 K and 77 K at gate voltages of 20 V, 40 V, and 60 V. The transfer characteristics in the linear regime as well as in the saturation regime are shown in Fig. 2 (a) and (b) respectively.

Temperature Dependence of the Charge

175

Fig. 1. The output characteristics of the OFET device shown at different gate voltages (a) at 300 K. (b) at 77 K.

Fig. 2. Transfer characteristics of the C60 FET at different temperatures. (a) In the linear regime for VD = 2V and (b) in the saturation regime for VD = 60V.

Without taking into account the contact resistance the drain current (ID) can be described in the linear regime as function of VG by the following equation.

ID

V D = co n stt .

=

W L

μ F E C i V D (V G − V th . )

(4 )

where W and L are the width and length of transistor channel, Ci the capacitance of the dielectric layer and Vth the threshold voltage. In the linear regime, the applied gate field is much larger than the in-plane drift field, which results in an approximately uniform density of charge carriers in the active channel. Using equation (4), μFE can be evaluated from the local slopes of the transfer characterization (see Fig. 2(a). The obtained μFE in the linear regime is plotted as a function of VG in Fig. 3 for different temperatures.

176

M. Ullah et al.

Fig. 3. (a) µFE in the linear regime as a function of carrier density and the VG at different T. (b) Arrhenius plot of field effect mobility in linear regime at gate voltages of 5 –11V.

This set of data can be converted into an Arrhenius plot as shown in Fig. 3(b) for gate voltages between 5 and 11 V. The mobility data can now be analyzed by using the MNR described in equation (3). From the slopes of the fit to data points activation energies for the mobility can be calculated for each gate voltage. The results are summarized in Fig. 4(a). The Arrhenius plot of Fig. 3(b) can be fitted by two different lines for each VG depending on the temperature range taken into account. Therefore we obtained two EA for each VG depending on the temperature range. The explanation of different EA in the different ranges of T is unclear, although it has also been observed previously in sexithiophene (6T) and octothiophene (8T) [2].

Fig. 4. (a) VG dependent EA evaluated using the data from Fig. 4 (a) in two different ranges of T. (b) Plot of µo verses EA using the data from Fig. 3 (b).

Following the MNR we can deduce from the common intersect of all the fits in Fig. 3(b) the µMN = 2.45 cm2V–1s–1 and the TMN = 480 K which corresponds to an EMN = 1/ (kB TMN) = 41 meV. Finally the extrapolation of the fitting lines in Fig. 3(b) for infinite temperature gives the prefactor µo as defined in the equation (1) for each EA. The data are summarized in Fig. 4(b).

Temperature Dependence of the Charge

177

Such an observation of strong VG and T dependent EA can be explained using the multi trap and release model which assumes a semiconductor with Fermi level closer to the band edge and upon applying VG the Fermi level moves through the distribution of band tail states. As a result EA is reduced as the density of injected charge carriers are increased above the mobility edge [4, 6–11].

4.

Conclusions

Organic field effect transistors were fabricated based on C60 layers. The temperature dependence of µFE was investigated for different VG. μFE is found to be thermally activated with activation energies strongly dependent on the applied VG. Upon extrapolation of the data in the Arrhenius plot to infinite temperature in the Meyer-Neldel formalism, TMN = 480 K and µMN = 2.45 cm2V–1s–1 can be extracted. The observed temperature dependence of µFE can be explained by empirical MNR, which is based on the assumption of an exponential density of states distribution. Acknowledgements. The work was supported by Austrian Science Foundation (FWF) within the National Research Network (NFN) “Interface controlled and Functionalized Organic Films”.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Th. B. Singh, N. S. Sariciftci, H. Yang, L. yang, B. Plochberger and H. Sitter, Appl. Phys. Lett. 90, 231512 (2007). G. Horowitz, R. Hajlaoui, R. Bourguiga, M. Hajlaoui, Synth. Met. 101, 401–404 (1999). J. Paloheimo and H. Isotalo, Synth. Met. 55, 3185 (1993). P. Stallinga and H. L. Gomes, Organic Electronics. 7, 529–599 (2006). Th. B. Singh, N. Marjanovic, G. J. Matt, S. Günes, N. S. Sariciftci, A. Montaigne Ramil, A. Andreev, H. Sitter, R. Schwödiauer and S. Bauer, Organic Elec. 6, 105–110 (2005) G. Horowitz, R. Hajlaoui and P. Delannoy J. Phys. III France 5, 355–371 (1995) E. j. Meijer, M. Matters, P. T. Herwig, D. M. de Leeuw and T. M. Klapwijk, Appl. Phys. Lett. 23, 76 (2000). P. Stallinga, H. L. Gomes, F. Biscarini, M. Muriga and D. M. de Leeuw, J. Appl. Phys. 9, 96 (2004). R. J. Chesterfield, J. C. Mackeen, C. R. Newman and C. D. Frisbie, J. Appl. Phys. 11, 95 (2004). P. Stallinga, H. L. Gomes, Organic Electronics 6, 137–141 (2005). G. Horowitz and M. E. Hajlaoui Adv. Mater. 12, 14, 1046–1050 (2000). M. C. J. M. Vissenberg and M. Matters, Phys. Rev. B 57, 12964 (1998). R. Metselaar and G. Oversluizen, J. Solid State Chemistry, 55, 320–326 (1984). W. Meyer and H. Neldel, Z. Tech. 18, 588 (1937) J. C. Wang and Y. F. Chen, Appl. Phys. Lett. 73, 948 (1998).

The Influence of Chain Orientation in the Electric Behaviour of Polymer Diodes Marta Ramos and Helder Barbosa Departamento de Física, Universidade do Minho, Campus de Gualtar, 4710-057 Braga, Portugal E-mail: [email protected] Abstract. Recently some experimental results have showed that the spatial alignment of conjugated polymer chains on nanometre length scales can influence the behaviour of polymer-based electronic devices, such as light-emitting diodes, field effect transistors, and photovoltaic cells. The effects of chain orientation at electrode-polymer interfaces on the charge injection process and charge mobility through the polymer layer are not well understood. In this work we use a generalized dynamical Monte Carlo method to study the influence of different polymer chain orientation relative to the electrodes surface on the electric behaviour of single-layer polymer diode, namely density current and charge density.

1.

Introduction

The use of conjugated polymers in electronic and optoelectronic applications goes from light-emitting diodes (LEDs) [1], to field-effect transistors (FETs) [2], and photovoltaic (PV) cells [3], among others. In all these devices, the morphology of the active polymer layer can critically influence the device behaviour [4,5]. The folding and bending of the polymer chains, that can be seen at the nanometre scale as a connection of conjugated segments with varied lengths, and their position and orientation with respect to neighbouring chains and/or the substrate, in which they are deposit, influences the electric properties of these polymer-based devices. However, from the experimental results [6], it is not possible to know the relation between the orientations of the polymer strands and the device electric behaviour. Most of the polymer films are prepared using the spin-cast technique that leads to polymer strands taking a preferential orientation parallel to the electrodes [7]. However, polymer chains tend to interpenetrate each other and there is the possibility of appearing domains within the polymer layer having perpendicular [8] and random [9] orientations relative to the electrodes surface. In this work, we study the influence of the orientation of the conjugated polymer segments on the electric behaviour of single-carrier polymer diodes, using a computational model based on a generalized dynamical Monte Carlo method, which includes explicitly the nanostructure of the polymer layer and the molecular properties of the polymer as input parameters.

2.

Device Model and Simulation Method

In order to build the nanostructure of a polymer diode with a specific morphology, we placed straight conjugated segments of poly(p-phenylenevenylene) (PPV)

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_29, © Springer-Verlag Berlin Heidelberg 2009

179

180

M. Ramos and H. Barbosa

(considered as rigid rods) randomly in the gap between two planar electrodes, separated 100 nm from each other, with their axis oriented parallel, perpendicular and randomly relative to the electrodes surface. The minimum distances (0.650 nm) allowed between the polymer strands and between those and the electrodes, well as the mean value (7 monomers) of a Gaussian distribution of strand lengths, were taken from previous theoretical and experimental results reported elsewhere [8,10]. In this model we consider that charge injection/collection from/by the electrodes and intermolecular charge transport within the polymer network occur by hopping with a frequency given by [10]:

wij = w0 × cosθ × exp( −

rij − r0 r0

Δε ij ⎧ ), for Δε ij > 0 ⎪exp( − )×⎨ k BT ⎪ 1, for Δε ij < 0 ⎩

(1)

where the first term represents the attempt-to-escape frequency and the following terms represent the influence of the direction of the local electric field (which is the sum of the applied electric field, the field due to charge distribution within the polymer network and the field due to electrode polarization), the hopping distance and the energy barrier height (which depends on the ionization potential and electron affinity of the involved polymer strands and their bias voltage) on the hopping process, respectively. Only the process with the highest hopping probability takes place. If the local electric field is higher than the field needed to move the injected charge along the polymer strand an intramolecular charge transport is also considered. A detailed description of the injection/collection and transport processes for ohmic contacts can be found in ref. [10,11] and the molecular properties used in this work as input parameters were taken from ref. [12]. Our dynamic model is based on the first reaction method (FRM) [13], where a queue of increasing time steps associated to the occurrence of the all electronic processes in the polymer diode is used to follow the time evolution of the charges in the device. For each electronic process there is a waiting time given by:

τ =−

ln( x) wij

(2)

where wij is the hopping frequency associated to an electronic process, being x a random number uniformly distributed between 0 and 1. At each computer iteration the electronic process with the smallest time of occurrence takes place and removed from the queue. This time of occurrence is then subtracted to waiting times, and new enable events are inserted in the queue.

The Influence of Chain Orientation in the Electric Behaviour of Polymer Diodes

3.

181

Results and Discussion

Using the three-dimensional model described above, we performed computational simulations of single-carrier (electrons or holes) injection and transport in polymer diodes for applied electric fields ranging from 0.3 MV/cm to 0.7 MV/cm. Since the results obtained for electrons and holes are similar, we just present here the results for electrons. Figure 1 shows the time evolution of current density for all strand orientations consider in this work. The variations in time of current density are due to stochastic time dependence of all electronic processes. An increase in the applied electric field leads to an increase of the current density especially for the parallel and random polymer strand orientations relative to the electrodes. The differences in the current density between the parallel, perpendicular and random orientations are due to the charge transport along the polymer network and the charge distribution within that network, both affecting charge injection. For the polymer layer with parallel morphology, since the number of hopping positions in a neighbouring strand is the largest on and the charge hopping occurs mainly in the same direction of the applied electric field, the charge transport is limited by the energetic disorder. Therefore, an increase in the electric field reduces the energetic barrier for hopping between neighbouring strands and as a result electrons can easily percolate along the polymer network reducing the effect of charge distribution within the polymer layer on the injection process. In the case of the polymer strands oriented perpendicularly to the electrodes surface, the processes of charge injection and charge transport are strongly dependent on the morphology of the polymer layer. First, since all the strands are perpendicular

Fig. 1. Variation of the current density with time for polymer diodes with parallel, perpendicular and random chain orientations and for applied electric fields of 0.3 MV/cm (straight line), 0.5 MV/cm (dash line) and 0.7 MV/cm (dot line).

182

M. Ramos, and H. Barbosa

to the electrodes surface, just the monomers near the electrodes are possible positions for charge injection. Second, since all the applied electric fields are smaller than the threshold for intramolecular charge mobility (1.55 MV/cm for electrons and 2.00 MV/cm for holes) [14], injected charges in any polymer strand moves along it towards its centre, which is the position energetically more favourable. Since the hopping distance to the neighbour strands favoured by the applied electric field increases, the hopping process takes longer. As a consequence, the injected charges will stay longer near the injection electrode, limiting charge injection by changing the internal electric field. The polymer network with a random orientation of the strands, exhibits an intermediary behaviour between the parallel and the perpendicular morphologies, for the same applied electric field. Figure 2 shows the change in time of charge density inside the polymer layer, for all the three morphologies considered in this work. The fact the charge transport along the network is more difficult for the polymer with the perpendicular morphology allows that charges stay longer inside the polymer network increasing charge density. The opposite behaviour is predicted for the polymer layer with all strands oriented parallel to the electrodes. For all the polymer morphologies, an increase in the applied electric field leads to an increase in charge density due to an increase of charge injection.

Fig. 2. Variation of the charge density with time for polymer diodes with parallel, perpendicular and random chain orientations and for applied electric fields of 0.3 MV/cm (straight line), 0.5 MV/cm (dash line) and 0.7 MV/cm (dot line).

The Influence of Chain Orientation in the Electric Behaviour of Polymer Diodes

4.

183

Conclusions

Although our model mimics in a simple way both the nanostructure and the electronic processes involved in a single-carrier polymer diode, the values for current and charge densities obtained are comparable to those measured experimentally. Besides, it can give some insights on the effect of polymer morphology on device performance that is impossible to obtain from experiments because in a real polymer film all polymer strand orientations considered in this work, as well as the presence of physical/chemical defects and the effects of both electrode/polymer interfaces contribute to the overall current density of the device. From our results, it is clear that when the polymer segments are packed parallel to the electrodes surface there is an increase in intermolecular charge transport compared to random and perpendicular orientations which leads to an increase of current density for the same applied electric field. Acknowledgements. This work is part of the research projects POCTI/CTM/41574/ 2001 and CONC-REEQ/443/EEI/2005, approved by the Portuguese Foundation for Science and Technology (FCT) and support by the European Community Fund FEDER. One of us (H.M.C.B.) is also indebted to FCT for financial support under PhD grant Nº SFRH/BD/22143/2005.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14

B.J. Schwartz, Annu. Rev. Phys. Chem. 54, 141, 2003. C. Tanase, E.J. Meijer, P.W.M. Blom, D.M. de Leeuw, Phys. Rev. Lett. 91, 21, 2003. H. Hoppe, N.S. Sariciftci, Journal of Materials Chemistry 16, 1, 2006. T.L. Benanti, D. Venkataraman, Photosynth. Res. 87, 1, 2006. M. Jaiswal, R. Menon, Polym. Int. 55, 12, 2006. J. Liu, Y.J. Shi, L.P. Ma, Y. Yang, J. Appl. Phys. 88, 2, 2000. C.Y. Yang, F. Hide, M.A. Diaz-Garcia, A.J. Heeger, Y. Cao, Polymer 39, 11, 1998. B.G. Sumpter, P. Kumar, A. Mehta, M.D. Barnes, W.A. Shelton, R.J. Harrison, J. Phys. Chem. B 109, 16, 2005. J. Kim, J. Lee, C.W. Han, N.Y. Lee, I.J. Chung, Appl. Phys. Lett. 82, 24, 2003. M.M.D. Ramos, H.M.G. Correia, J. Phys.: Condens. Matter 18, 2006. H.M.C. Barbosa, M.M.D. Ramos, Plasma Process. Polym. 4, 2007. A.M. Stoneham, M.M.D. Ramos, A.M. Almeida, H.M.G. Correia, R.M. Ribeiro, H. Ness, A.J. Fisher, J. Phys.: Condes. Matter 14, 42, 2002. J.J. Lukkien, J.P.L. Segers, P.A.J. Hilbers, R.J. Gelten, A.P.J. Jansen, Phys. Rev. E 58, 2, 1998. A.M. Almeida, M.M.D. Ramos, H.G. Correia, Comput. Mater. Sci. 27, 1–2, 2003.

Interface Modification of Pentacene OFET Gate Dielectrics Ján Jakabovič1, Jaroslav Kováč1, Rudolf Srnánek1, Jaroslav Kováč jr.1, Michal Sokolský2, Július Cirák2, Daniel Haško3, Roland Resel4 and Egbert Zojer 4 1

Department of Microelectronics, Slovak University of Technology, Ilkovičova 3, 812 19 Bratislava, Slovakia E-mail: [email protected] 2 Department of Physics, Slovak University of Technology, Ilkovičova 3, 812 19 Bratislava, Slovakia 3 International Laser Center, Ilkovičova 3, 812 19 Bratislava, Slovakia 4 Institute of Solid State Physics, Graz University of Technology, Petergasse 16, A-8010, Graz, Austria Abstract. Pentacene organic field effect transistors (OFETs) electrical and structural properties have already been analysed from the point of view of different gate dielectric and growth conditions utilization. The AFM and micro Raman investigations show that the first organic monolayer at the pentacene/dielectric interface are essential determinants of carrier transport phenomena and achievable drain current of pentacene OFETs.

1.

Introduction

Pentacene is one of the leading candidates of the many organic materials available, for use in current organic field-effect transistor (OFET) architectures, because of its excellent electrical characteristics and its resistance to atmospheric oxygen [1]. Manipulating the semiconductor/dielectric interfacial properties via optimising the gate dielectric can substantially enhance OFET performance [2,3]. Parylene’s superior electrical insulation characteristics make it an excellent solution for the OFETs gate dielectric. In this work we present pentacene/gate dielectric interface modification with thin parylene layers and self-assembled monolayer (SAM) of diacetylene formed by the Langmuir-Blodgett method.

2.

Experimental Methods

In order to understand the pentacene thin films growth on different dielectric materials we studied the growth of pentacene molecules on three different substrate materials SiO2, SiO2 covered with diacetylene SAM formed by the LangmuirBlodgett method and SiO2 covered with thin parylene layer. Pentacene bottom gate top contact OFET structures were prepared on silicon substrate (gate) covered with thermal grown 40 nm thick silicon dioxide with these different dielectric layers to compare their electrical properties. The polymer parylene C has been chosen as a gate insulator material, which forms transparent pinhole−free conformal coatings with excellent dielectric. High purity dichloro-di-para-xylylene was used for chemical H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_30, © Springer-Verlag Berlin Heidelberg 2009

185

186

J. Jakabovič et al.

vapour deposition process to create high quality thin parylene C layers 20 and 40nm thick. The pentacene films were prepared from commercially available material (Acros Organic) with 98% purity. Pentacene was deposited by thermal evaporation at substrate temperature of 30, 50 and 70°C and deposition rate 0,2 – 0,5 Å/s. For preparation of the OFET with the top electrodes structure, gold source and drain electrodes with different channel length (15, 25, 45 μm) and 2000 μm channel width were deposited thermally thru the shadow mask on 40 nm pentacene layer.

3.

Results and Discussion

0

0

-20

-20

-40

VG= -20 V, Pentacene, TGrowth SiO2 + A174-Silane, 70°C

-60

SiO2 + SAM,

70°C

SiO2 + Parylene C, 70°C

-80

SiO2 + Parylene C, 30°C

0 -2 -4 -6 -8 -10 -12 -14 -16 -18 -20

-60 -80

-100

-100 -120

VG [ V ]

-40

IDS [mA]

IDS [μ A]

Figure 1 show output characteristics of the organic pentacene OFETs prepared on different SiO2, SAM and parylene dielectric layers at constant gate voltage VG = −20 V. A significant increase of the source–drain current were measured for OFETs prepared on parylene layers in comparing with devices prepared on SiO2 based interface layers. The highest value of the drain current was reached for pentacene layer prepared at 30°C on parylene dielectric layer. The measured output characteristics in dependence on gate voltage are shown in Fig. 2, corresponding to charge mobility of 0.15 cm2V–1s–1. The analysis of the measured results shows the pentacene/dielectric interface has considerably influence on the carrier transport phenomena. This was confirmed by AFM investigations of the thin pentacene layers prepared at different growth temperature 30 and 70°C on parylene layers. As shown in Fig. 3, 4 the pentacene layer growing at 30°C shows better crystalline structure formation. For the thick 40 nm pentacene layer prepared at 30°C on parylene dielectric the AFM image (Fig. 5) shows creation of two crystalline structures (thin film and bulk). This was confirmed with microRaman spectroscopy measurements (Fig. 6) where the bands 1154 cm–1 and 1158 cm–1 corresponds to formation of two crystalline structures. The band at 1154 cm–1 is highly related to carrier transport of pentacene films, because it reflects the coupling between molecules of pentacene. Similar results were obtained by including of PMMA layer under pentacene layer [4].

-120 -25

-20

-15

-10

-5

0

VDS [ V ]

Fig. 1. OFET output characteristic for different dielectrics at constant VG.

Pentacene (40 nm) on 40 nmParylene grown at 30°C

-30

-25

-20

-15

-10

-5

0

VDS [V]

Fig. 2. OFET output characteristics for pentacene grown on parylene dielectric.

Interface Modification of Pentacene OFET Gate Dielectrics

a)

187

b)

Fig. 3. AFM image (3 × 3 μm) of thin 3 nm pentacene layer grown at 30°C on parylene layer.

Fig. 4. AFM image (3 × 3 μm) of thin 3 nm pentacene layer grown at 70°C on parylene layer. o

2,5

2,0

2,0

1,5

1,5

1,0 0,5

1,0 0,5

Raman intensity [a.u.]

40 nm pentacene grown @ 30 C on 20 nm parylene C

3

A1154 / A1158 = 1.19 -1

1154 cm

-1

1158 cm

2

1 1150

1175

1200 -1

Raman shift [cm ]

Fig. 5. 3D AFM image of 40 nm pentacene layer grown at 30°C on parylene layer.

Fig. 6. 40 nm pentacene grown at 30°C on parylene Raman spectrum.

In summary, the measured results confirmed that the gate dielectric layer at the pentacene/dielectric interface is essential origins, which considerably influenced the carrier transport phenomena and achievable drain current of pentacene OFETs. Acknowledgements. This work was supported by grant of Slovak Research and Development Agency No. APVV-0290-06 and VEGA-1/3108/06, 1/0742/08 projects.

References 1. 2. 3. 4.

D.J. Gundlach, Y.Y. Lin, T.N. Jackson, S.F. Nelson and D.G. Schlom, in IEEE Electron Device Letters, 18, 87, 1997 Ch. Pannemann, T. Diekmann and U. Hilleringmann, in Microelectronic Engineering, 67–68, 845, 2003 J.H. Park, C.H. Kang, Y.J. Kim, Y.S. Lee and J.S. Choi, in Materials Science and Engineering C, 24, 27, 2004 H. Cheng, Y. Mai, W. Chou, L. Cheng, in Appl. Phys. Lett., 90, 171926, 2007

Negative Differential Resistance in C60 Diodes Philipp Stadler1, Anita Fuchsbauer1, Günther Hesser2, Thomas Fromherz3, Gebhard J. Matt1, Helmut Neugebauer1 and Serdar N. Sariciftci1 1

Linz Institute for Organic Solar Cells, Physical Chemistry, Johannes Kepler University Linz, Altenbergerstraße 69, 4040, Austria E-mail: [email protected] 2 Technische Service Einheit (TSE), Johannes Kepler University Linz, Altenbergerstraße 69, 4040 Linz, Austria E-mail: [email protected] 3 Institute for Semiconductor and Solid State Physics, Johannes Kepler University Linz, Altenbergerstraße 69, 4040, Austria E-mail: [email protected] Abstract. Morphology studies and current-voltage (IV) measurements of C60 thin film diodes in the temperature range of 300–4.2 K are presented. For defined evaporation parameters orientation domains along the growth direction are demonstrated by cross section transmission electron microscopy. From the electrical characterization the fullerene diodes exhibit space charge limited currents which follow a power law dependency. At current densities above 100 mA cm–2 and temperatures below 200 K reversible voltage instabilities (S-shape IV characteristics, negative differential resistance) arise. The instabilities are similar to charge transport effects in amorphous inorganic semiconductors.

1.

Introduction

Fullerene and its derivatives are dominantly used as electron acceptor molecules in conducting polymer matrices for solar cells. The investigation of the transport properties of these materials are therefore of importance. Here we concentrate on the properties of C60 films processed by thermal evaporation in sandwich type diodes [1]. The high purity and reproducibility in thin films grown by evaporation allow the study of both, the transport properties as well as the morphology of these materials. The nature of charge transport in fullerenes and in amorphous or polycrystalline inorganic semiconductors is similar. In both cases charge injection follows the space charge limited current (SCLC) theory. With C60, a reversible voltage instability is observed at low temperatures corresponding to similar effects in amorphous inorganics [2,3]. The origin of the voltage breakdown is designated to charge trapping next to the electrodes, followed by the formation of conductive filaments in the bulk C60 layer.

2.

Experimental Methods

Sandwich-type diodes were fabricated using 70 nm thick poly(3,4ethylenedioxythiophene/poly(styrene-sulfonate) (PEDOT:PSS) spin coated film on ITO glass as hole injection electrode. 300 nm C60 films were grown on top by

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_31, © Springer-Verlag Berlin Heidelberg 2009

189

190

P. Stadler et al.

thermal evaporation. The substrate temperature was kept constant at 140°C during the deposition process. The morphology studies were perfomed by transmission electron microscopy (TEM). For the transport measurements we used 50 nm Chromium layers as top contact. All electrical characterizations were performed in a Helium flow cryostat (Oxford Industries) in a temperature range between 300 K to 4.2 K. The voltage drop between the PEDOT:PSS and the Chromium electrode was measured by applying a certain current density to the diode.

3.

Results

The device structure and the energy levels of the fullerene diode are shown in Figure 1. The estimated built-in voltage between the Fermi level of the Chromium (–4.5 eV) and the Fermi level of PEDOT:PSS (–5.2 eV) is about 0.7 V. The low energy of the HOMO of the C60 results in an energy barrier of around 0.6 eV for the hole injection which is comparatively higher than for the electron injection.

Fig. 1. Device sandwich – type structure and band diagrams of the C60 diode.

The electrical characterization was performed in the high injecton regime (forward direction, Chromium as cathode). For gaining the effective voltage drop the built-in potential (Ueff = Umeasured – UBI) is subtracted from the measured voltage drop in the device. In Figure 2 the IV characteristics at 3 temperatures are shown.

Fig. 2. Double logarithmic IV characteristics at 300 K, 250 K and 200 K.

Negative Differential Resistance in C60 Diodes

191

The IV behaviour is described with a formation of a space charge upon the dominant electron-injection (space charge limited currents, SCLC). The current density follows a power law dependence, described with the Mott-Gournier formula in (1)

9 V2 n J = ⋅ Θ ⋅ ε ⋅ μ ⋅ 3 and Θ = free 8 L ninjected

(1)

where µ is the mobility, L the thickness of the diode and Θ the ratio of free to injected charge carriers. In ideal case (no traps) the ratio is close to 1 and the exponent is 2. In C60 the injection process is more complex due to presence of shallow and deep traps. We see experimentally a higher exponent (Fig. 2), since the factor Θ is also voltage-dependent. As in many organic systems, also with evaporated C60 the morphology influences the transport properties. The cross section TEM studies verify high crystallinity of C60 and lateral faceted crystallite domains along the growth direction (Fig. 3). Trapping is present at crystallite grain boundaries and the free carrier concentration is dependent on the temperature and the morpholgy [4,5].

Fig. 3. (a) Dark-field analysis of diode cross section and zoom in (c) shows similar orientation domains in growth direction. (b) Diffraction pattern of C60 crystallite exhibits fcc hexagonal structure. (d) High resolution dark field shows orientation domain of the molecules.

The power law dependency is fulfilled when operating with current densities around 100 mA cm–2. The exponent increases by decreasing the temperature, as seen in Figure 2. At current densities above 100 mA cm–2 and temperatures below 200 K deviations from the SCLC behaviour are observed. By applying a higher current density the measured voltage decreases showing a negative-differential resistance (NDR) seen in Figure 4.

192

P. Stadler et al.

Fig. 4. Negative differential resistance regime at 4.2 K showing a reversible voltage breakdown. Inset: Voltage drop measured time-resolved in negative differential regime.

The IV foward and reverse scan exhibits a well pronounced negative differential resistance regime at 4.2 K above 150 mA cm–2. Consequently we can exclude thermal effects as origin for the voltage breakdown. The inset plot in Figure 4 shows the measured voltage drop time resolved. In the first 20 ms of the current pulse the voltage remains at the value corresponding to a SCLC behaviour, then the voltage drops to a constant level until the end of the 250 ms current pulse.

4.

Conclusions

It is known that transport in C60 is influenced by the purity and crystallinity of the material. In this work the evaporation parameters are controlled in order to grow high quality films on the PEDOT:PSS anode. From the morphology studies in figure 3 the authors state that crystallites with similar oriented domains are formed in the growth direction. These films exhibit in the electrical transport characterization space charge limited (SCL) currents, which follow a power law dependency as expected from the theory. The current voltage characteristics from the diode correspond to the SCL current with presence of traps. At temperatures below 200 K and current densities above 100 mA cm–2 a reversible voltage instability arises. In analogy to amorphous inorganic semiconductors, this phenomena is described with the formation of conductive filaments in the bulk of the C60 film (injection over a non-uniform injection area). Since the high conductivity at low temperatures cannot be explained with a thermal activation process, we propose a band transport model and an existance of a mobility edge [6]. Acknowledgements. We gratefully acknowledge the Austrian FWF Project and the Austrian NFN Project for support.

Negative Differential Resistance in C60 Diodes

193

References 1 2 3 4 5 6

H. Sitter, A. Andreev and N. S. Sariciftci, Mol. Cryst. Liq. Cryst., 385, 171 (2002). G. J. Matt, T, Fromherz and N. S. Sariciftci, Appl. Phys. Lett., 84, 1570 (2004). N. F. Mott, Contemp. Phys., 10, 125–138 (1969). K. Rikitake, T. Akiyama and W. Takashima, Synth. Met., 86, 2357 (1997). G. Krakow, N. M Rivera, and J. J. Cuomo, Appl. Phys. A., 56, 185–192 (1995). G. J. Matt, T. Fromherz, H. Neugebauer and N. S. Sariciftci, Electronic Properties of Novel Nanostructures, 530 (2005).

Performance and Transport Properties of Phthalocyanine:Fullerene Organic Solar Cells M. Rusu1, J. Gasiorowski1, S. Wiesner1, D. Keiper2, N. Meyer2, M. Heuken2, K. Fostiropoulos1 and M.Ch. Lux-Steiner 1 1

Department SE2, Hahn-Meitner-Institut Berlin, Glienicker Strasse 100, 14109 Berlin, Germany E-mail: [email protected] 2 AIXTRON AG, Kackertstr. 15-17, 52072 Aachen, Germany E-mail: [email protected] Abstract. Copper phthalocyanine (CuPc)–fullerene (C60) photovoltaic cells are produced by organic vapour phase deposition reaching efficiencies of 3%. The electronic transport properties of the devices are investigated as a function of the CuPc:C60 absorber blend layer composition and its preparation temperature. The analysis of the transport properties of the devices employs the one-diode model. It is shown that the dominant recombination process takes place at the donor–acceptor interfaces of the CuPc and C60 absorber domains. The activation energy of recombination is related to the effective band gap of the blend layer.

1.

Introduction

Organic solar cell (OSC) efficiencies of 5% were recently achieved on indium tin oxide (ITO)/CuPc/CuPc:C60/C60/bathocuproine (BCP)/Al photovoltaic (PV) devices employing donor (D) copper-phtalocyanine (CuPc) and acceptor (A) fullerene C60 materials [1]. However, little is known about how the device performance and electrical properties are influenced by the composition and preparation conditions of the CuPc:C60 blend layer. In this contribution, we discuss the impact that CuPc:C60 absorber composition and its preparation temperature has on device PV parameters as well as on electrical and transport properties. ITO/3,4-polyethylene-dioxythiophene: polystyrenesulfonate (PEDOT:PSS)/CuPc:C60/Mg/Ag OSCs are investigated.

2.

Experimental

ITO/PEDOT:PSS/CuPc:C60/Mg/Ag organic solar cells were fabricated on ITO (5 Ω/square sheet resistance)-coated glass substrates. After solvent cleaning, the ITO/glass substrates were spin-coated by a PEDOT:PSS layer and immediately transferred into the deposition chamber. A 70 nm-thick CuPc:C60 blend layer was prepared by organic vapour phase deposition (OVPD®) [2, 3]. The Mg/Ag back contacts were deposited by thermal evaporation in high vacuum (p ∼ 10–7 mbar) on non-air-exposed absorber surfaces. The device preparation details can be found elsewhere [4]. The compositional and substrate temperature (Tsubstrate) investigations are carried out on type A and B devices with nonoptimised and optimised contacts, respectively. H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_32, © Springer-Verlag Berlin Heidelberg 2009

195

196

M. Rusu et al.

PV parameters and power conversion efficiency (Eff), fill factor (FF), opencircuit voltage (Voc) and short-circuit current density (Jsc) and were determined from J-V measurements performed under standard conditions (AM1.5; 100 mW/cm2; 25°C). Measurements of the J-V characteristics as a function of temperature (JV-T) and illumination were performed in an evacuated custom made N2-cooled cryostat. A set of neutral density filters (SCHOTT) served for adjusting the illumination, ranging from 5 × 10–4 mW/cm2 to 100 mW/cm2. The J-V curves were analysed by the one-diode model developed for inorganic thin film solar cells [5].

3.

Results and Discussion

Performance and Diode Parameters. For finding the optimum absorber composition, organic solar cells of the type A with various [CuPc]:[C60] ratios were prepared at a non-optimised substrate temperature. The Eff behaviour with the absorber composition (Fig. 1a) is dominated by the behaviour of the Jsc and FF [4]. In the compositional range investigated, 0.2 ≤ [CuPc]/([CuPc] + [C60]) ≤ 0.8, the devices’ Voc is almost constant taking values of about 400 mV (not shown). An efficiency of 1.6% is achieved in Fig. 1a at a [CuPc]:[C60] composition of about 1:1 (by weight) [4]. The diode quality factor, n, of the devices decreases from 2.62.7 at the either minimum of CuPc or C60 content to ∼1.8 at 1:1 absorber composition. At the same time, the OSCs’ series resistance decreases almost symmetrically from 0.34–0.38 Ω × cm2 at a content of 17% of either CuPc or C60 to ∼0.2 Ω × cm2 at an identical content of CuPc and C60. The influence of the preparation temperature on the devices characteristics was investigated on both types A and B samples prepared at the optimum [CuPc]: [C60] = 1:1donor-to-acceptor ratio. The solar cells’ efficiency (Fig. 1b) is mainly dominated by the temperature behavior of Jsc and FF (not shown). For both types of samples a maximum efficiency is achieved at Tsubstrate ∼150°C. However, the Btype samples with the optimized contacts show by up to 3 times higher efficiencies compared to A-type devices. At the optimum substrate temperature the best B-type OSC shows the following PV parameters: Eff = (3.0 ± 0.3)%, Voc = 470 mV, Jsc = 15.1 mA/cm2 and FF = 0.42.

a)

3.0

Efficiency (%)

Efficiency (%)

1.5

1.0

0.5

non-optimised contacts optimised contacts

b)

2.5 2.0 1.5 1.0 0.5

0.0

0.0

0.2

0.4

0.6

[CuPc]/([CuPc]+[C60])

0.8

130

140

150

160

170

180

Substrate temperature (°C)

190

Fig. 1. Efficiency of ITO/PEDOT:PSS/CuPc:C60/Mg/Ag OSCs as a function of CuPc:C60 blend layer (a) composition and (b) preparation temperature.

non-optimised contacts optimised contacts

2.5

2.0

1.5 130 140 150 160 170 180 190

Substarte temperature (°C)

2

a)

Series resistance Rs (Ω cm )

Diode quality factor n (-)

Performance and Transport Properties of Phthalocyanine

1.2

b)

197

non-optimised contacts optimised contacts

1.0 0.8 0.6 0.4 0.2

130

140

150

160

170

180

Substrate temperature (°C)

190

Fig. 2. (a) Diode quality factor and (b) series resistance of the ITO/PEDOT:PSS/ CuPc:C60/ Mg/Ag OSCs as a function of preparation temperature. The doted lines are guides to the eye.

The diode quality factor n of type A devices decreases as a function of substrate temperature in Figure 2a from 2.4–2.6 at Tsubstrate = 131°C or Tsubstrate = 187°C to 1.5 at Tsubstrate = 151°C. The behavior of the series resistance (Rs) with temperature is shown in Fig. 2b. The minimum Rs = 0.27 Ω × cm2 is achieved for type B OSCs at Tsubstrate = 148°C. The enhancement of the devices PV and diode parameters with the temperature up to ∼150°C can be explained by (i) an improved separation of the CuPc and C60 donor and acceptor materials in an interpenetrated absorber network, (ii) enhanced crystalline perfection of the CuPc domains [4] and therefore improved transport properties, i.e., better collection efficiency of photogenerated carriers at the respective electrode. Alteration of the photoelectrical parameters at higher temperatures can be attributed to the potential degradation of the PEDOT buffer layer. Transport Properties of the type B devices with optimized contacts were investigated on a solar cell prepared at optimum conditions – Tsubstarte = 150°C and [CuPc]:[C60] = 1:1. The diode parameters, i.e., n and J0 – the saturation current density, were determined from the JV-T and JscVoc-T curves in the dark and under illumination, respectively. In the measurement temperature range 240–320 K, constant n values of about 1.5–1.6 and 1.2–1.3 were determined in the dark and under illumination, respectively. The ideality factor increases at low temperatures of about 200 K to ∼2 and ∼1.8 in the dark and under illumination, correspondingly. Thus, in the whole temperature region investigated the recombination process is thermally activated (n < 2). The activation energy (Ea) of charge carrier recombination was determined according to Ref. 5 from n × lnJ0 = f(1/T) curves. It is found that Ea changes from 0.79 eV under illumination to 1.02 eV in the dark. These values are much lower than the band gaps of C60 – Eg-C60 = 2.3 eV [6] or CuPc – Eg-CuPc = 1.7 eV [7], whereas they are very close to the energy difference of ∼0.7 eV between the CuPc-HOMO (HOMO: highest occupied molecular orbital) and C60LUMO (LUMO: lowest unoccupied molecular orbital) levels [8] defined as the effective band gap of the CuPc:C60 blend absorber. Thus, the current transport in this type of OSCs is limited by the recombination of charge carriers at interfaces of CuPc and C60 domains. Because the same energy difference |HOMOCuPc – LUMOC60|

198

M. Rusu et al.

determines the Voc of the solar cell [9], we conclude that Ea and Voc are correlated [10], as for inorganic solar cells.

4.

Conclusions

In summary, up to 3%-efficient ITO/PEDOT:PSS/CuPc:C60/Mg/Ag OSCs with an OVPD-prepared CuPc:C60 absorber were demonstrated. The composition of the donor:acceptor blend layer at an optimum preparation temperature has a strong impact on the devices’ short-circuit currents, diode quality factors as well as their series resistances, while the open circuit voltages are almost unaffected in the compositional range investigated. From investigations of the devices photovoltaic and photoelectric properties it can be concluded that a substrate temperature of ∼150°C provides formation of an absorber with improved photoelectrical and transport properties. The analysis of the OSCs transport properties shows that the main recombination occurs at interfaces of CuPc and C60 domains. The recombination process is thermally activated in a large temperature range 200–320 K, the activation energy being of the order of the absorber effective band gap. Acknowledgements. This work is supported by the German Ministry of Environment (Contract No. 0329927A). OVPD® technology has been exclusively licensed to AIXTRON from Universal Display Corporation (UDC), Ewing, N.J. USA for equipment manufacture. OVPD® technology is based on an invention by Prof. S. R. Forrest et al. at Princeton University, USA, which was exclusively licensed to UDC. AIXTRON and UDC have jointly developed and qualified OVPD® pre-production equipment.

References 1 2

F. Yang, M. Shtein, and S. R. Forrest, J. Appl. Phys. 98, 014906, 2005. M. Heuken and N. Meyer, in Organic Electronics (Weinheim, Ger.), Edited by H. Klauk, ISBN-10: 3-527-31264-1, ISBN-13: 978-3-527-31264-1-Wiley-VCH. 3 M. Rusu, S. Wiesner, T. Mete, H. Blei, N. Meyer, M. Heuken, M.Ch. Lux-Steiner, and K. Fostiropoulos, Renewable Energy 33, 254, 2008. 4 M. Rusu, J. Gasiorowski, S. Wiesner, N. Meyer, M. Heuken, K. Fostiropoulos, and M.Ch. Lux-Steiner, Thin Solid Films (2007), doi:10.1016/j.tsf.2007.12.004. 5 V. Nadenau, U. Rau, and A. Jasenek, J. Appl. Phys. 87, 584, 2000. 6 R. W. Lof, M. A. van Veenendaal, B. Koopmans, H. T. Jonkman, and G. A. Sawatzky, Phys. Rev. Lett. 68, 3924, 1992. 7 I. G. Hill, A. Kahn, Z. G. Soos, and R. A. Pascal, Chem. Phys. Lett. 327, 181, 2000. 8 P. Peumans and S. R. Forrest, Appl. Phys. Lett. 79, 126, 2001. 9 D. Chirvase, Z. Chiguvare, M. Knipper, J. Parisi, V. Dyakonov, and J.-K. Hummelen, J. Appl. Phys. 93, 3376, 2003. 10 M. Rusu, J. Strotmann, M. Vogel, M.Ch. Lux-Steiner, and Fostiropoulos, Appl. Phys. Lett. 90, 153511, 2007.

Organic Transistors Based on Molecular and Polymeric Dielectric Materials Antonio Facchetti*, Sara DiBenedetto, Choongik Kim, Tobin J. Marks* Department of Chemistry and the Materials Research Center, Northwestern University, 2145 Sheridan Road, Evanston, IL 60208, USA Abstract. The design and synthesis of new molecular synthons for vapor-phase selfassembled nanodieletrics and silane crosslinkers for crosslinked polymer blend dielectrics is described. These dielectric films exhibit excellent dielectric properties with tunable thicknesses and capacitance values. These new gate dielectric materials are integrated into thin-film transistors based both p- and n-type organic semiconductors.

1.

Introduction

Small molecule and polymeric dielectric thin films with tuned permittivities (ε) have recently found applications in organic field effect transistors (OFETs), where replacing the traditional gate dielectric (SiO2) with high capacitance materials enables lower OFET operating voltages, reduced power consumption, and improved device performance.1 The gate dielectric capacitance Ci in a OFET can be described by a parallel plate capacitor, where Ci = εεo/d (d is the dielectric thickness and εo is the permittivity of vacuum). To achieve large capacitance values it is necessary to reduce the thickness and/or increase the permittivity of the dielectric film. This approach creates an interesting materials challenge not only because small molecule and polymeric materials with large permittivities (i.e. π-conjugation) are usually conductors, but also because as the insulator film thickness is reduced the leakage current increases to a level detrimental for device operation.2 In this work two new dielectric classes for OFETs, v-SAND and CPB, are described.

2.

Experimental Methods

v-SAND Materials. Compounds 1 and 2 (see Fig. 1 for structure) were synthesized following the general procedure reported elsewhere and were characterized by conventional analytical/spectroscopic techniques. v-SAND films are fabricated according to the scheme reported in Fig. 1, and are composed of a bottom organic (1 or 2) layer and capped with a top SiOx matrix. Films of 1 and 2 (F1 and F2, respectively) were vapor deposited at a rate of 0.2 Å/sec onto solvent cleaned heavily doped n+-Si substrates maintained at room temperature and under high vacuum (10–6 Torr). To deposit the inorganic SiOx capping layer, F1 and F2 were exposed to trichlorodisiloxane (Cl6Si2O, 96%, Aldrich) vapors under N2 for 20 min and annealed under vacuum (20 Torr) for 4 hr at 110°C. Film thicknesses were measured by X-ray reflectivity (XRR). Metal-insulator-semiconductor (MIS) 2 devices were fabricated by depositing Au electrodes (200 × 200 μm ) directly on

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_33, © Springer-Verlag Berlin Heidelberg 2009

199

200

A. Facchetti et al.

Fig. 1. Chemical structures of 1 and 2 and OFET device fabrication steps: (a) vapor deposition of 1 or 2, (b) exposure of film 1 or 2 to Cl6Si2O vapors to complete the v-SAND structure, (c) vapor deposition of pentacene, and (d) vapor deposition of Au electrodes through a shadow mask.

all dielectric films. Bottom-gate top-contact OFETs were fabricated by vapor deposition of pentacene on the organic-inorganic dielectrics, and the I-V characteristics were analyzed according to typical procedures. CPB Materials. Various types of silane crosslinkers were employed to fabricate crosslinked polymer blend (CPB) dielectrics in this study (Fig. 4). The reactivity of each crosslinker was tested via in situ NMR kinetic studies. The CPB dielectrics were fabricated on various substrates using mixture of polymer and crosslinker solution via spin-coating and gravure-printing. Organic semiconductors and source/ drain electrodes were vacuum-deposited to complete the OFET device. Dielectric and OFET properties were measured under vacuum and ambient as described previously.

3.

Results and Discussion

Vapor-Phase Self-Assembled Nanodielectric (v-SAND) Materials. Dielectric materials composed of molecular components are ideal candidates for OFETs because small molecules are on the length scale of nanometers, may be of tunable electronic structure, can be processed on flexible substrates, and are compatible with organic semiconductors.3 Previously our group demonstrated the application of molecular self-assembled nanodielectrics (SANDs) as gate dielectrics. SAND films are fabricated via a layer-by-layer solution phase self-assembly of σ-π silane molecular precursors to form hybrid organic-inorganic multilayers for OFETs.3 An interesting question is whether similar hybrid films based on highly π-polarizable components can be fabricated from the vapor phase in a process that can overcome the synthesis of sophisticated silane precursors necessary for solution phase selfassembly. In this contribution, we demonstrate SAND-like organic-inorganic films as gate dielectrics in OFETs fabricated entirely through the vapor phase (v-SAND). The new v-SAND π-conjugated building blocks 1 and 2 were designed to form robust self-ordered thin films via head-to-tail intramolecular hydrogen bonding

Organic Transistors Based on Molecular

201

(Fig. 1). In addition the π-conjugated donor-acceptor organic functionalities should enhance electronic polarizabilities, which is attractive to achieve high permittivity materials. Pentacene OFETs based on v-SANDs exhibit excellent performance at low operating voltages. As in the case of solution processed SANDs4, a substantial leakage current (J) reduction in is observed (10 – 10–2 → 10–5 – 10–7 A/cm2 at 2 V) after the capping layer is deposited (F1, F2 → F1-cap, F2-cap). The leakage current reduction indicates that the capping material functions as an efficient electron tunneling barrier.4

Fig. 2. A. Leakage current vs. voltage plots of F1 (red dotted line) and F2 (blue dotted line) and F1-cap (red solid line) and F2-cap (blue solid line). B. XRR spectra for F1 (red) and F2 (blue), and C. XRR spectra F1-cap (red) and F2-cap (blue).

Because of the large J of F1 and F2, meaningful capacitances are only measured for F1-cap and F2-cap. MIS devices of both F1-cap and F2-cap exhibit large capacitances of approximately 400 nF/cm2 at 2 V. Permittivities of the v-SAND organic components can be estimated by modeling F1-cap and F2-cap dielectric layers as three parallel plate capacitors in series (Si native oxide + F1 or F2 + SiOx). From the XRR data (Fig. 2b and 2c), the capping layer and the organic layer thicknesses are ~3.6 nm and ~5.9 nm, respectively. From the accumulation regime capacitances of F1-cap (400 nF/cm2) and F2-cap (390 nF/cm2), ε values of ~11 and ~9 are found for F1 and F2, respectively (using dnative oxide = 1.5 nm and εnative oxide = εcapping layer = 3.9). The larger ε of F1 vs. F2 is anticipated by the INDO(SOS) computations of the molecular polarizabilities in the context of the Clausius-Mossotti equation, and can be explained by the larger acidity of 1 upon intramolecular hydrogen bonding. Using the above capacitances and channel dimensions: L = 200 μm and W = 5000 μm, these OFETs exhibit excellent performance with hole mobilities of 1.9 ± 0.3 cm2/Vs and 2.4 ± 0.3 cm2/Vs for the F1-cap and F2-cap based devices, respectively, Ion/Ioff ~ 105, and VT ~ 1 V. Note that output plot I-V curves at each gate voltage cross at the origin indicating very low gate leakage current during device operation (Fig. 3). These results demonstrate that vapor phase deposition can probe and utilize the dielectric properties of high-ε molecular layers without compromising the insulating properties of the corresponding hybrid films. To understand the origin of the large carrier mobilities, AFM images of the pentacene films on top of v-SANDs were collected (Fig. 3). Clearly, very large pentacene crystal grains are formed (~2 – 3 μm), which is comparable to and larger than those in other high-mobility OFETs.

202

A. Facchetti et al.

Fig. 3. Transfer (A and D) and output (B and E) plots for pentacene OFETs based on v-SANDs of F1-cap (top) and F2-cap (bottom). AFM images and XRD spectra of the corresponding 50 nm thick pentacene films grown on F1-cap (C) and F2-cap (F).

Crosslinked Polymer Blend (CPB) Dielectric Materials. The CPB materials consist of chemically bonding poly(vinyl phenol) (PVP) with silane crosslinkers (Fig. 4). Previous CPB dielectrics using chlorosilane-based crosslinkers afforded relatively rough dielectric surfaces due to high reactivity of the crosslinking reagents and unoptimized crosslinking reaction.5 To control and optimized the crosslinking conditions affording robust films with smooth surface morphology for improved OFET performance, various solvents and polymer-crosslinker concentratin ratio were tested. Table 1 shows the film deposition condition describing optimization of <20 nm-thick spin-coated CPB films. As shown, moderately reactive C6OAc and EGOAc afford relatively smoother morphology and lower leakage current densities compared to highly reactive C6Cl and C6NMe2. The least reactive C6OMe does not form robust films. Similar results with moderately reactive crosslinkers showing the best dielectric properties were obtained for spin-coated thick and gravure-printed films.6 The OFETs on CPB films show excellent device performance comparable to conventional 300 nm thick SiO2. n

Source

Drain

PVP

Semiconductor Crosslinked Polymer Blend (CPB) n+-Si (Gate)

OH

Cl Si Cl Cl

Cl Cl Si Cl

C6Cl

(Me)2N Si (Me)2N N(Me)2

N(Me)2 N(Me)2 N(Me)2

Si

C6NMe2

+

OAc OMe OMe OAc MeO Si Si OMe OAc MeO OMe C6OAc C6OMe OAc OAc AcO O Si Si O OAc AcO OAc EGOAc

AcO Si AcO OAc

Si

Fig. 4. Schematic of top-contact/bottom-gate OFET device and structures of the CPB precursors (polymer and silane crosslinkers) employed in this study.

Organic Transistors Based on Molecular

203

Table 1. Spin-coated CPB thin film (<20 nm) deposition condition and pentacene-based TFT device performance (mobility and on/off ratio) – polymer/crosslinker concentration ratio (mg/ml:mg/ml), solvent, film thickness (D, nm), RMS roughness (ρ, nm), leakage current density at an electric field of 2 MV/cm (J, A/cm2), mobility (μ, cm2/Vs), and Current On/off Ratio (Ion:Ioff). Crosslinker

Ratio

Solvent

D

ρ

J

μ

Ion:Io ff

C6NMe2 C6Cl C6OAc EGOAc C6OMe

4.

4:6 6:6 4:6 4:6 4:4

EtOAc Dioxane EtOAc EtOAc EtOAc

17 14 13 14 15

0.7–0.8 0.5–0.6 0.3–0.4 0.2–0.3 0.3–0.4

–7

~8 × 10 8–70 × 10–8 4–70 × 10–8 5–50 × 10–8 >1 × 10–5

0.12 0.18 0.35 0.37 –

105 104 104 104 –

Conclusions

We have demonstrated a facile vapor-phase method for fabricating self-assembled nanodielectrics (v-SANDs) from molecular precursors exhibiting large capacitances and good insulating properties. Pentacene OFETs based on F1-cap and F2-cap exhibit very large mobilities. Furthermore, we have shown that blending of commercially-available polymer with appropriate silane crosslinking agents affords robust, smooth, adherent, and pin-hole free dielectric materials which are readily deposited from solution and afford well-performing pentacene TFTs. These results show that robust, low-leakage polymer dielectric films are accessible and that the TFT devices implementing these polymer dielectrics in solutionprocessed fabrication methodologies offer opportunities for printing application. Acknowledgements. This work was supported by the NSF MRSEC program (DMR-0520513) at the Materials Research Center of Northwestern University, by the ONR MURI Program (N00014-02-1-0909), and Polyera.

References 1 2 3 4 5 6

H. Klauk. Organic Electronics: Materials, Manufacturing, and Application, WILEYVCH, Weinheim, Germany, 2006. A. Facchetti, M.-H. Yoon, T. J. Marks, Adv. Mater. 17, 1705, 2005. M. -H. Yoon, A. Facchetti, and T. J. Marks, Proc. Natl. Acad. Sci. 102, 4678, 2005. S. A. Dibenedetto, D. Frattarelli, M. A. Ratner, A. Facchetti, T. J. Marks, J. Am. Chem. Soc. 130, 7528, 2008. M. -H. Yoon, H. Yan, A. Facchetti, T. J. Marks, J. Am. Chem. Soc. 127, 10388, 2005. C. Kim, Z. Wang, H. -J. Choi, Y. -G. Ha, A. Facchetti, T. J. Marks. J. Am. Chem. Soc. 130, 6867, 2008.

springer proceedings in physics 129

springer proceedings in physics 113 Theoretical and Numerical Unsaturated Soil Mechanics Editor: T. Schanz 114 Advances in Medical Engineering Editor: T.M. Buzug 115 X-Ray Lasers 2006 Proceedings of the 10th International Conference, August 20–25, 2006, Berlin, Germany Editors: P.V. Nickles, K.A. Janulewicz 116 Lasers in the Conservation of Artworks LACONA VI Proceedings, Vienna, Austria, Sept. 21–25, 2005 Editors: J. Nimmrichter, W. Kautek, M. Schreiner 117 Advances in Turbulence XI Proceedings of the 11th EUROMECH European Turbulence Conference, June 25–28, 2007, Porto, Portugal Editors: J.M.L.M. Palma and A. Silva Lopes 118 The Standard Model and Beyond Proceedings of the 2nd International Summer School in High Energy Physics, M¯gla, 25–30 September 2006 Editors: M. Serin, T. Aliev, N.K. Pak 119 Narrow Gap Semiconductors 2007 Proceedings of the 13th International Conference, 8–12 July, 2007, Guildford, UK Editors: B. Murdin, S. Clowes 120 Microscopy of Semiconducting Materials 2007 Proceedings of the 15th Conference, 2–5 April 2007, Cambridge, UK Editors: A.G. Cullis, P.A. Midgley

122 Advances in Nanoscale Magnetism Proceedings of the International Conference on Nanoscale Magnetism ICNM-2007, June 25–29, Istanbul, Turkey Editors: B. Aktas, F. Mikailov 123 Computer Simulation Studies in Condensed-Matter Physics XIX Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 124 EKC2008 Proceedings of the EU-Korea Conference on Science and Technology Editor: S.-D. Yoo 125 Computer Simulation Studies in Condensed-Matter Physics XX Editors: D.P. Landau, S.P. Lewis, and H.-B. Sch¨uttler 126 Vibration Problems ICOVP 2007 Editors: E. Inan, D. Sengupta, M.M. Banerjee, B. Mukhopadhyay, and H. Demiray 127 Physics and Engineering of New Materials Editors: D.T. Cat, A. Pucci, and K.R. Wandelt 128 Ultrasonic Wave Propagation in Non Homogeneous Media Editors: A. Leger, M. Deschamps 129 Interface Controlled Organic Thin Films Editors: H.-G. Rubahn, H. Sitter, G. Horowitz, K. Al-Shamery

121 Time Domain Methods in Electrodynamics A Tribute to Wolfgang J. R. Hoefer Editors: P. Russer, U. Siart

Volumes 90–112 are listed at the end of the book.

¨ Horst-Gunter Rubahn Helmut Sitter Giles Horowitz Katharina Al-Shamery Editors

Interface Controlled Organic Thin Films With 122 Figures

123

Prof. Dr. Horst-Günter Rubahn

Prof. Dr. Helmut Sitter

University of Southern Denmark NanoSYD Mads Clausen Institute Alsion 2, 6400 Sønderborg Denmark [email protected]

Universität Linz Institut für Experimentalphysik Abteilung Festkörperphysik 4040 Linz Austria [email protected]

Prof. Dr. Giles Horowitz

Prof. Dr. Katharina Al-Shamery

Université Paris VII Institut Topologie et de Dynamique Systèmes (ITODYS) 1 rue Guy-de-la-Brosse 75005 Paris France [email protected]

Universität Oldenburg Institut für Reine und Angewandte Chemie 26111 Oldenburg Germany [email protected]

ISSN 0930-8989 ISBN 978-3-540-95929-8 e-ISBN 978-3-540-95930-4 DOI 10.1007/978-3-540-95930-4 Springer Dordrecht Heidelberg London New York Library of Congress Control Number: “PCN applied for” © Springer-Verlag Berlin Heidelberg 2009 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: SPi Publisher Services Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Preface Organic electronics is a scientific and technological field that has witnessed an enormous world-wide effort both in basic scientific research as well as in industrial development within the last decades. It is becoming increasingly clear that, if devices based on organic materials are ever going to have a significant relevance beyond being a cheap replacement for inorganic semiconductors, there will be a need to understand interface formation, film growth and functionality. A control of these aspects will allow the realisation of totally new device concepts exploiting the vast flexibility inherent in organic chemistry. The field of devicerelevant “semiconducting” organic materials has many parallels to that of inorganic semiconductors. However, the versatility of organic molecules comes at the cost of higher complexity of the materials. This rules out a 1:1 transfer of concepts established within inorganic semiconductor research to the world of organics, and makes work on organic semiconductors particularly challenging. On a world-wide scale, investigations of organic thin films focus on three main areas with different aims and with a fruitful mixture of applied and basic research: (1) the development and production of devices, (2) thin film characterization and more recently, after recognizing the importance of molecular level control (3) surface and interface science. Linking these branches together creates new synergies and has led and leads to a significant advance in the field of organic semiconductors. Eventually it will result in the development of the necessary tools for tuning device properties on a nanoscopic level. In the last 10 to 15 years a large amount of investigations of devices have been performed with a big range of active organic materials. This work has mapped out the classes of materials that proof useful for single molecule, oligomeric/molecular films and plastic electronics. In this symposium we focused on oligomeric/molecular films, because the control of molecular structures and interfaces provides unprecedently highly defined systems. This in turn allows one to study basic physics and at the same time enables one to find the important parameters necessary to improve organic devices. The E-MRS symposium conceived to bring together the leading groups, which work in the field of growth and characterisation of organic films and devices and focus them on the fabrication and characterisation of highly ordered functional organic films. The wide range of expertise of the contributing groups allowed the combination of different methodologies and aspects of physics, chemistry, and materials science for the design and understanding of well-defined organic structures. In total we received 148 contributions to the symposium in the form of invited talks, oral presentations and posters. Out of them the reviewers selected a representative amount of papers to be published in the proceedings. The main topics discussed at the symposium are reflected in the headlines of the chapters in the proceedings. Introductory review papers based on invited talks given at the symposium are followed by contributed papers. The highlights of the oral and poster presentations contributing to the same topic are summarized in the same chapter. The editors would like to thank the sponsors of the E-MRS symposium, especially the ‘Fonds der Chemischen Industrie’.

vi

Preface

In technical finishing the book we would like to thank Ms. Zora Milde for her extraordinary help in mastering the handling of all the electronic documents.

Sønderborg, Linz, Paris, Oldenburg January 2009

H.-G. Rubahn H. Sitter G. Horowitz K. Al-Shamery

Contents Preface .................................................................................................................... V A Thin Film Growth............................................................................................ 1 1

Toward an Ab-initio Description of Organic Thin Film Growth ............... 3 P. Puschnig, D. Nabok, and C. Ambrosch-Draxl

2

Organic Nano Fibres from PPTPP .............................................................. 11 F. Balzer, M. Schiek, A. Lützeu, and H.G. Rubahn

3

α-Sexithiophene Films Grown on Cu(110)–(2x1)O: From Monolayer to Multilayers ................................................................................................. 19 M. Oehzelt,, S. Berkebile, G. Koller, T. Haber, M. Koini, O. Werzer, R. Resel, and M.G. Ramsey

4

Para-Sexiphenyl Layers Grown on Light Sensitive Polymer Substrates ....................................................................................................... 23 G. Hernandez-Sosa, C. Simbrunner, T. Höfler, A. Moser, O. Werzer, B. Kunert, G. Trimmel, W. Kern, R. Resel and H. Sitter

5

Thermal Desorption of Organic Molecules................................................. 29 A. Winkler

6

Crystalline Stages of Rubrene Films Probed by Raman Spectroscopy................................................................................................... 37 B.A. Paez, Sh. Abd-Al-Baqi, G.H. Sosa, A. Andreev, C. Winder, F. Padinger, C. Simbrunner, and H. Sitter

7

Rubrene Thin Film Characteristics on Mica .............................................. 43 Sh.M. Abd Al-Baqi, G. Henandez-Sosa, H. Sitter, B. Th. Singh, Ph. Stadler, N.S. Sariciftci

8

Structural Properties of Rubrene Thin Films Grown on Mica ........................................................................................................ ...49 T. Djuric, H.-G. Flesch, M. Koini, Sh.M. Abd Al-Baqi, H. Sitter,and R. Resel

vii

viii

9

Contents

Rubrene on Mica: From The Early Growth Stage To Late Crystallization................................................................................................ 55 G. Hlawacek, S. Abd-al Baqi, X. Ming He, H. Sitter, and C. Teichert

10 β-Sheeted Amyloid Fibril Based Structures for Hybrid Nanoobjects on Solid Surfaces......................................................................................... ..61 V. Bukauskas, V. Strazdienė, A. Šetkus, S. Bružytė, V. Časaitė,and R. Meškys 11 Characteristics of Vacuum Deposited Sucrose Thin Films .................... ..67 F. Ungureanu, D. Predoi, R.V. Ghita, R.A. Vatasescu-Balcan,and M. Costache 12 Electropolymerization of Polypyrrole Films in Aqueous Solution with Side-Coupler Agent to Hydrophobic Groups.................................. ..73 H.M. Alfaro-López, J.R. Aguilar-Hernandez, A. Garcia-Borquez, M.A. Hernandez-Perez,and G.S. Contreras-Puente. 13 Surface Modification of Polymer Powders by a Far Cold Remote Nitrogen Plasma in Fluidized Bed............................................................. ..79 L. Aiche, H. Vergnes, B. Despax, B. Caussat,and H. Caquineau 14 Features of Polytetrafluoroethylene Coating Growth on Activated Surfaces from Gas Phase .......................................................................... ..85 A.A. Rogachev, S. Tamulevičius, A.V. Rogachev, I. Prosycevas,and M. Andrulevičius 15 Modification of Amorphous Carbon Film Surfaces by Thermal Grafting of Alkene Molecules.................................................................... ..91 H. Sabbah, A. Zebda, S. Ababou-Girard, B. Fabre, S. Députier, A. Perrin, M. Guilloux-Viry,and F. Solal, C. Godet 16 DNA-Controlled Assemblage of Ag Nanoparticles on Solid Surfaces .. ..95 V. Bukauskas, A. Šetkus, I. Šimkienė, J. Sabataitytė, A. Kindurys, and A. Rėza, J. Babonas

Contents

ix

17 Characterization of Self Assembled Monolayer Formation of 11 - Mercaptoundecanoic Acid on Gold Surfaces .............................. 101 J. Stettner, P. Frank, T. Griesser, G. Trimmel, R. Schennach, R. Resel, and A. Winkler 18 SAMs of 11-MUA Grown on Polycrystalline Au-foils by Physical Vapor Deposition in UHV...................................................................................... 107 P. Frank, F. Nussbacher, J. Stettner,and A. Winkler 19 Photoreactive Self Assembled Monolayers for Tuning the Surface Polarity ........................................................................................................ 113 T. Griesser, A. Track, G. Koller, M. Ramsey W. Kern, and G. Trimmel B Traps and Defects ....................................................................................... 119 20 Spectroscopy of Defects in Epitaxially Grown Para-sexiphenyl Nanostructures............................................................................................ 121 A. Kadashchuk, S. Schols, Yu. Skryshevski, I. Beynik, C. Teichert , G. Hernandez-Sosa, H. Sitter, A. Andreev, P. Frank, and A. Winkler 21 Magnetoresistance in Poly (3-hexyl thiophene) Based Diodes and Bulk Heterojunction Solar Cells........................................................ 127 S. Majumdar, H. S. Majumdar, H. Aarnio, R. Laiho, and R. Österbacka 22 Evolution of the Bipolaron Structure in Oligo-diacetylene Films: A Semiempirical Study .............................................................................. 133 M. Ottonelli, G. Musso, and G. Dellepiane C Energy Level Alignment and Charge Transfer ....................................... 139 23 Molecular Orientation Dependence of the Ionization Energy of Pentacene in Thin Films ........................................................................ 141 G. Heimel and N. Koch 24 Charge transfer and Polarization Screening at Organic/Metal Interfaces: Single Crystalline versus Polycrystalline Gold .................... 147 H. Peisert, D. Kolacyak, A. Petershans,and T. Chassé

x

Contents

25 Sensing Infrared Light with an Organic/Inorganic Hetero-junction.... 153 Gebhard J. Matt, Thomas Fromherz , Guillaume Goncalves, Christoph Lungenschmied, Dieter Meissner, and Serdar N. Sariciftci D Advanced Characterization Methods ....................................................... 159 26 Ultrafast Confocal Microscope for Functional Imaging of Organic Thin Films ................................................................................................... 161 Dario Polli, Michele Celebrano, Jenny Clark, Giulia Grancini, Tersilla Virgili, Guglielmo Lanzani, and Giulio Cerullo 27 Growth and Desorption Kinetics of Sexiphenyl Needles: An In-situ AFM/PEEM Study ..................................................................................... 167 Alexander J. Fleming, Svetlozar Surnev, Falko P. Netzer, and Michael G. Ramsey E

Organic Devices .......................................................................................... 171

28 Temperature Dependence of the Charge Transport in a C60 based Organic Field Effect Transistor ................................................................ 173 Mujeeb Ullah, Th.B. Singh, G.J. Matt, C. Simbruner, G. Hernandz-Sosa, S.N. Sariciftci,and H. Sitter 29 The Influence of Chain Orientation in the Electric Behaviour of Polymer Diodes........................................................................................................... 179 Marta Ramos and Helder Barbosa 30 Interface Modification of Pentacene Ofet Gate Dielectrics .................... 185 Ján Jakabovič, Jaroslav Kováč, Rudolf Srnánek, Jaroslav Kováč jr., Michal Sokolský, Július Cirák, Daniel Haško, Roland Resel, and Egbert Zojer 31 Negative Differential Resistance in C60 Diodes ........................................ 189 Philipp Stadler, Anita Fuchsbauer, Günther Hesser, Thomas Fromherz, Gebhard J. Matt, Helmut Neugebauer, and Serdar N. Sariciftci 32 Performance and Transport Properties of Phthalocyanine: Fullerene Organic Solar Cells ................................................................... 195 M. Rusu, J. Gasiorowski, S. Wiesner, D. Keiper, N. Meyer, M. Heuken, K. Fostiropoulos, and M.Ch. Lux-Steiner1

Contents

xi

33 Organic Transistors Based on Molecular and Polymeric Dielectric Materials...................................................................................................... 199 A. Facchetti, S. DiBenedetto, C. Kim,and T.J. Marks 34 Morphology of the Metal-Organic Semiconductor Contacts: the Role of Substrate Surface Treatment ................................................ 205 A.Petrović, E. Pavlica, and G. Bratina 35 Molecular Interactions Between Alcohols and Metal Phthalocyanine Thin Films for Optical Gas Sensor Applications..................................... 211 S. Uttiya, S. Kladsomboon, O. Chamlek, W. Suwannet, T. Osotchan, T. Kerdcharoen, M. Brinkmann, and S. Pratontep 36 Organic Thin-Film Transistors with Enhanced Sensing Capabilities.................................................................................................. 217 M. Daniela Angione, F. Marinelli, A. Dell’Aquila, A. Luzio , B. Pignataro and L. Torsi 37 Photoelectric Properties of Microrelief Organic/Inorganic Semiconductor Heterojunctions................................................................ 225 N.L. Dmitruk, O.Yu. Borkovskaya, D.O. Naumenko, I.B. Mamontova, N.V. Kotova, O.S. Lytvyn, and Ya.I. Vertsimakha List of Contributors........................................................................................... 229

Morphology of the Metal-Organic Semiconductor Contacts: The Role of Substrate Surface Treatment Andraž Petrović, Egon Pavlica and Gvido Bratina Laboratory for epitaxy and nanostructures, University of Nova Gorica, Vipavska 11c, 5270 Ajdovščina, Slovenia Email: gvido.bratina @p-ng.si

Abstract. We have systematically investigated the role of SiO2 surface treatment on pentacene morphology at the metal/pentacene interface prior to vacuum deposition of the pentacene layer onto OTFT structures. The substrates included as grown SiO2 and SiO2 surface treated by self-assembled monolayer (SAM) of hexamethyldislazane (HMDS). The resulting OTFT’s were investigated in situ, during growth of pentacene layer by measurements of current-voltage characteristics, and ex situ by an atomic force microscopy (AFM). Our results show that the effective field-effect mobility of pentacene decreases with decreasing growth rate. In addition, for low growth rates, we observed an existence of discontinuous pentacene coverage at the metal-pentacene interface. We associate the decrease in mobility to this morphological feature. We have found that HMDS treatment results in a reduced areal density of extended structural defects. However, at the lowest growth rates even HMDS treatment can not promote wetting of the metallic contacts with pentacene in order to close the gap between pentacene and the metallic contact.

1.

Introduction

Organic thin film transistors (OTFT’s) based on pentacene are likely to become an important component in advanced electronic applications [1–6] due to the demonstrated high charge carrier mobility. A typical OTFT includes thin organic semiconductor (OS) layer deposited between two metallic contacts on a suitable substrate. Most of the present OTFT’s are fabricated by vacuum evaporation of pentacene onto SiO2-covered highly-doped Si substrates that act as a gate electrode. The morphology of a pentacene layer evaporated onto SiO2 is characterized by grains whose shape and size depend on processing parameters such as deposition rate [7,8], substrate temperature [7,8] and surface treatment of the insulator [9]. Morphology of the organic semiconductor (OS) layer was found to affect the electronic transport OTFTs. Horowitz using octithiophene (8T) [10] and Di Carlo et al. [11] using pentacene, both report on decreasing field-effect mobility with increasing density of grain boundaries. On the other hand, if SiO2 is treated by self-assembled monolayer (SAM) prior to the OS layer growth, the field-effect mobility was found to increase despite the increasing density of grain boundaries compared to not treated samples [9,12].

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_34, © Springer-Verlag Berlin Heidelberg 2009

205

206

A. Petrović, E. Pavlica and G. Bratina

In addition to the morphological features of the pentacene layer, the performance of an OTFT is influenced by the microscopic interface environment at the interface between the pentacene layer and a source and drain metallic contact. The electronic parameters of the interface may give rise to an increased contact resistance. Therefore it is important to understand the relationship between the chemical/structural characteristics of the OS/metal interface and charge carrier transport in OTFT. For example the difference in mobility between the top-contact and bottom-contact OTFT was associated to the different morphology of the pentacene layer near the metallic contacts [13]. In this work we have investigated the influence of pentacene film morphology near the metallic contacts on electric charge transport in OTFTs. The morphology of pentacene layers was controlled by the flux of pentacene molecules. For lower growth rates we observe a region near the metal contact where no pentacene layers were observed. We associated this feature to lower “effective” field-effect mobility measured in our experiment. Additionally, we investigated the morphology of pentacene layers in the vicinity of the metallic contacts for treated and untreated substrates. We ascribed the difference in morphology to the treatment of SiO2 surface with HMDS.

2.

Experimental Methods

In our experiments we have used bottom-gate, bottom-contact OTFT’s Interdigitated Au source/drain structures were fabricated by optical lithography on SiO2-covered p++ silicon substrates. The device channel length was L = 5 μm and the channel width was W = 2.3 mm. Two types of substrates were employed in our experiments: one group of the samples included substrates treated by HMDS prior to the pentacene deposition, in the other group of the samples pentacene was grown directly onto SiO2. Pentacene was evaporated in a high-vacuum chamber, with the base pressure of 5 × 10–6 Pa. The deposition rate ranged from 0.05 nm/min up to 1.2 nm/min, and the substrate was at room temperature during pentacene growth. The pentacene layer thickness ranged from nominally 15 nm to nominally 18 nm, and was determined in situ by quartz thickness monitor. Electrical measurements during growth were performed using a Keithley 617 as a power supply and a Keithley 2400 as ammeter. The measurement of the currentvoltage (I-V) characteristic was initiated when the mechanical shutter blocking the sample surface off the pentacene molecules was removed. The gate voltage during an I-V sweep was –15 V, and the source-drain voltage ranged between +5 and –20 V. The time interval between successive sweeps varied between 1 and 8 s. In order to minimize the transistor stressing, the source-drain and the gate voltage was put to 0 V between successive sweeps. Experiments showed that extended transistor operation could result in a reduction of the source-drain current through the transistor [14] and in a shift of threshold voltage [15]. After the deposition of an active layer, the morphology was observed ex situ by an atomic force microscope (AFM) operating in air. In order to investigate in depth the eventual effects of microscopic interface environment on the performance of our OTFT’s we focused our AFM investigations to the regions in close proximity of the metallic contacts of the source and drain.

Morphology of the Metal-Organic Semiconductor Contacts

3.

207

Results and Discussion

Figure 1(a) shows the effective field-effect mobility plotted in logarithmic scale. The curves correspond to the nominal growth rates of 1.2 nm/min, 0.15 nm/min and 0.05 nm/min, top to bottom respectively. The effective field-effect mobility values were calculated from the I-V curves in the saturation regime [16]. After the onset of growth, at a pentacene layer thickness of (1.3 ± 0.2) nm we observe a sharp increase of mobility. After that the increase in mobility becomes slower. We note that the maximum value obtained for mobility depends on the pentacene growth rate (Fig. 1(a)). Assuming that a single-layer thickness of pentacene amounts to 1.5 nm [17], we may conclude that a connected current pathway between the source and the drain contact exists, even before a continuous pentacene layer is achieved. This is in accord to the results by Ruiz et al. [18], who studied the morphology of pentacene on SiO2 as a function of layer. Their results show that a second molecular layer starts to nucleate before a continuous first layer is formed. The formation of a continuous layer therefore proceeds by merging of the isolated islands.

Fig. 1. Evolution of an effective field-effect mobility as a function of pentacene layer thickness obtained in the samples fabricated by three different growth rate (1.2 nm/min, 0.15 nm/min and 0.05 nm/min, top to bottom) (a). Evolution of an effective field-effect mobility as a function of thickness obtained in the sample fabricated by a growth rate of 0.05 nm/min (b).

In Fig. 1(b) we present an effective field-effect mobility vs. pentacene layer thickness in the samples fabricated by the deposition rate 0.05 nm/min. We see that at such low growth rates the charge transport mechanisms evolve differently than at the elevated growth rates. The most prominent feature is marked plateau in mobility at the pentacene layer thickness range between 3 nm and 10 nm. The differences in the electronic transport are reflected also in the pentacene morphology near the metal/OS interface. In Fig. 2 we exemplify AFM scans of pentacene layers deposited on HMDStreated SiO2 using different growth rates. The scanned area shown is 4 × 4 μm2. The individual images correspond to the growth rates of 1.2 nm/min, 0.15 nm/min and 0.05 nm/min, Figs. 2a, 2b and 2c, respectively. A white region in Figs. 2a, 2b, and 2c corresponds to source metallic contact. The thickness of the pentacene layers is 18 nm, 17 nm and 15 nm in Figs. 2a, 2b and 2c, respectively. The thicknesses

208

A. Petrović, E. Pavlica and G. Bratina

Fig. 2. Atomic force micrographs of pentacene layer deposited on SiO2 treated with HMDS with three different deposition rates. The scan size is 4 × 4 μm2. (a) deposition rate 1.2 nm/min; (b) deposition rate 0.15 nm/min; (c) deposition rate 0.05 nm/min. Images were scanned in the vicinity of metallic source-drain contacts. The narrow white stripe indicates.

correspond to the maximum thicknesses used to determine mobility vs. thickness curves shown in Fig. 1. We see that the pentacene layer is characterized by grains, which size depends on the growth rate. For the growth rate of 1.2 nm/min the average grains size is 0.45 μm, for growth rate of 0.15 nm/min the average grains size is 0.78 μm, and for the growth rate of 0.05 nm/min the average grains size is 0.82 μm. This is what we would expect based on the deposition-diffusion-aggregation (DDA) model [19]. Similar behavior was also observed experimentally by Zuppiroli et al. [7] and is a signature of reduced surface mobility of the pentacene molecules as the flux of incoming molecules increases. We note that Fig. 2c shows a region near the metallic contact of the width between 10 nm and 120 nm, where no pentacene islands are present (indicated by arrows). This indicates that low growth rates may result in a discontinuous pentacene film near the metallic contacts. Electric charge transport across such interface is less efficient than in the case of homogeneous metal/pentacene interface. This is reflected by more than one order of magnitude lower effective field-effect mobility, observed in the samples fabricated by the growth rate of 0.05 nm/min (Fig. 1(a), bottommost curve). The existence of plateau in mobility vs. thickness curve (Fig. 1(b)) is likely to be a consequence of gradual closing of the uncovered region near the metallic contacts. When comparing the samples fabricated with the growth rate of 1.2 nm/min and the samples fabricated with the growth rate of 0.15 nm/min, the latter exhibits almost one order of magnitude lower field-effect mobility (Fig. 1(a)), despite the larger grain size (Fig. 2). We note that the microscopic interface environment near the metallic contacts exhibits no dewetting in both cases (Figs. 2(a) and 2(b)). The observed almost one-order-of-magnitude difference in the field-effect mobility therefore can not be explained by the morphological features near the metallic contacts. Similar results were obtained also in the experiment performed by Stein et al. [9], who treated SiO2 surface with SAM’s prior to the pentacene deposition. Their AFM investigation showed a decreasing field-effect mobility with increasing grain size, as observed in our experiment. We therefore submit that increased density of grain boundaries (smaller grain size) does not necessarily lead to a higher field effect mobility. the metallic contact. Area in image (c) indicated by arrows shows a region near the metallic contact of width between 10 nm and 120 nm, where no pentacene islands are present.

Morphology of the Metal-Organic Semiconductor Contacts

209

Fig. 3. Atomic force micrographs of pentacene layer deposited on SiO2 at a growth rate of 0.05 nm/min. The scan size is 4 × 4 μm2. The images correspond to pentacene layer thickness of 10 nm. Image 3(a) represents the sample where HMDS treatment was not applied prior to pentacene layer deposition. Black arrows indicate long continuous channel between metallic contact and pentacene layer where no pentacene islands are observed. White arrows indicate rifts in pentacene layer. Image 3(b) represents the sample where HMDS treatment was applied prior to pentacene layer deposition. Black arrows indicate irregularities at metal/ pentacene interface.

The effect of HMDS treatment on the pentacene layer morphology is presented in Fig. 3. When the surface was not treated with HMDS a strong evidence of pentacene dewetting of the metallic contact is observed in a form of a long continuous channel between the metallic contact and a pentacene layer (Fig. 3(a)). No pentacene islands are observed inside this channel. The width of the channel amounts up to 500 nm. When the surface was treated with HMDS we have also observed dewetting, albeit on a considerably smaller scale (Fig. 3(b)). In addition, samples without HMDS exhibited rifts in the pentacene layer even far from the metallic contacts (Fig. 3a, white arrows). We observed no such features on the samples treated with HMDS (Fig. 3(b)). We can therefore conclude that HMDS promotes coalescence between neighboring islands. We therefore submit that surface treatment improves the morphology of pentacene layers near the metallic contacts, resulting in a reported higher filed-effect mobility [9,12].

4.

Conclusions

We performed in-situ electrical characterization of pentacene-based organic thin film transistors, and examined ex situ the morphology of the pentacene channels. We observed a strong dependence of the effective field-effect mobility on the pentacene growth rate. Samples fabricated with higher growth rate exhibit higher field-effect mobility. The reduction in mobility with the reduction in growth rate is coupled in marked change in pentacene layer morphology. Samples fabricated by high growth rate exhibit higher effective field-effect mobility and smaller pentacene grain size. Samples fabricated by a low growth rate exhibit relatively

210

A. Petrović, E. Pavlica and G. Bratina

large grain size and lower effective field-effect mobility. This marked difference in effective field-effect mobility may originate from a relatively broad region of uncovered substrate near the source and drain contacts. Further, our results show that HMDS treatment improves the morphology of pentacene layers at the metal/pentacene interface, and may be responsible for higher reported filed-effect mobility of HMDS-treated OTFTs. Acknowledgements. We appreciate fruitful discussion with Rebernik Ribič and S. Stanič. The work was financed in part by the Ministry of higher education, science and technology under the research program P1-0040, and by the Center of excellence “Nanosciences and Nanotechnology”. We are in debt to L. Sorba and E. Di Fabrizio for assistance in lithographic process.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

C. D. Dimitrakopoulos, P. R. L. Malenfant, Adv. Mat., 14, 99 (2002). A. Facchetti, Mater. Today, 10, 28 (2007). Y. Y. Noh, D. Y. Kim, Solid-State Elect., 51, 1052 (2007). I. Manunza, A. Bonfiglio, Biosens. Bioelectron., 22, 2775 (2007). J. Jang, Mater. Today, 9, 46 (2006). G. S. Ryu, K. B. Choe, C. K. Song, Thin Solid Films, 514, 302 (2006). S. Pratontep, M. Brinkmann, F. Nüesch, L. Zuppiroli, Syn. Met., 146, 387 (2004). H. Yanagisawa, T. Tamaki, M. Nakamura, K. Kudo, Thin Solid Films, 464–465, 398 (2004). M. Shtein, J. Mapel, J. B. Benziger, S. R. Forrest, Appl. Phys. Lett., 81, 268 (2002). G. Horowitz, Adv. Mat., 12, 1046 (2000). A. Di Carlo, F. Piacenza, A. Bolognesi, B. Stadlober, H. Maresch H., Appl. Phys. Lett., 86, 263501 (2005). D. Knipp, R. A. Street, A. Völkel, J. Ho, J. Appl. Phys., 93, 347 (2003). I. Kymissis, C. D. Dimitrakopoulos, S. Purushothaman, IEEE Trans. Elect. Dev., 48, 1060 (2001). B. Fraboni, A. Matteucci, A. Cavallini, E. Orgiu, A. Bonfiglio, Appl. Phys. Lett., 89, 222112 (2006). S. Cipolloni, L. Mariucci, A. Valletta, D. Simeone, F. De Angelis, G. Fortunato, Thin Solid Films, 515, 7546 (2007). G. Horowitz, Adv. Mat., 10, 365 (1998). C. D. Dimitrakopoulos, A. R. Brown, A. Pomp, J. Appl. Phys., 80, 2501 (1996). R. Ruiz, B. Nickel, N. Koch, L. C. Feldman, R. F. Haglund, A. Kahn, G. Scoles, Phys. Rev. B, 67, 125406 (2003). B. A. -L. Barabasi and H. E. Stanley, Fractal Concepts in Surface Growth, Cambridge University Press, Cambridge (1995).

Molecular Interactions Between Alcohols and Metal Phthalocyanine Thin Films for Optical Gas Sensor Applications Sureeporn Uttiya1, Sumana Kladsomboon1, Onanong Chamlek1, Wiriya Suwannet1 , Tanakorn Osotchan1, Teerakiat Kerdcharoen1, Martin Brinkmann2 and Sirapat Pratontep3* 1

Center of Nanoscience and Nanotechnology and Department of Physics, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400, Thailand E-mail: [email protected] 2 Institut Charles Sadron, 23, rue du Loess, 67034 Strasbourg, France E-mail: [email protected] 3 Thai Microelectronics Center (TMEC-NECTEC), 51/4 Suwintawong Road, Muang, Chachoengsao 24000, THAILAND E-mail: [email protected] Abstract. Optically active organic gas sensors represent a promising molecular sensing device with low power consumption. We report experimental and computational investigations into the molecular interactions of metal phthalocyanine thin films with alcohol vapor. In the gas-sensing regime, the interactions of zinc phthalocyanine and alcohol molecules were studied by the Density Functional Theory (DFT) calculations, in comparison to the x-ray absorption spectroscopy. The DFT results reveal a reversible charge interaction mechanism between the zinc atom and the oxygen atom in the alcohol OH group, which corresponds to a shift in the x-ray absorption edge of the zinc atom. In the irreversible interaction regime, the effect of saturated alcohol vapor on spin-coated zinc phthalocyanine films was studied by the phase contrast microscopy, the optical absorption spectroscopy, and the transmission electron microscopy. Annealing the spin-coated films in saturated methanol vapor was found to induce an irreversible structural transformation from an amorphous to a crystalline phase, similar to the effect of a thermal annealing process. These crystallization processes of the zinc phthalocyanine films were also found to enhance their stability and alcohol sensing performance.

1.

Introduction

Gas sensors based on optically active properties of organic materials offer several key benefits over conventional metal oxide gas sensors, such as low power consumption, high gas specificity and simple fabrication processes [1]. Metal phthalocyanines and porphyrins are excellent examples of such gas-sensing materials, which have been demonstrated to detect various organic solvents and toxic gases [2,3]. Nevertheless, the fundamental understanding of the molecular mechanisms for optical gas sensors has been rarely investigated, particularly in predicting responses of organic sensors to various chemical vapors [4].

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4_35, © Springer-Verlag Berlin Heidelberg 2009

211

212

S. Uttiya et al.

Our previous work has shown that spin-coated films of zinc phthalocyanines exhibited detectable shifts in the optical absorption spectra during alcohol vapor exposure [5]. We here report an investigation of the molecular interactions between zinc phthalocyanines and alcohols by the x-ray absorption spectroscopy, the phase contrast optical microscopy and the transmission electron microscopy. The experimental results are also compared with our Density Functional Theory (DFT) calculations of the interactions between alcohol molecules and the zinc atom of the phthalocyanines [5].

2.

Experimental Methods

Zinc(II)-2,9,16,23-tetra-tert-butyl-29H,31H-Phthalocyanine (ZnTTBPc) was obtained from Sigma-Aldrich. The X-ray Absorption Spectroscopy (XAS) was performed at the Beamline 8 of the National Synchrotron Research Center (Thailand) [6]. The ZnTTBPc powder was packed in a small Pyrex container (1.5 × 0.6 × 0.2 cm3), sealed with polyimide tapes. To expose the films to alcohol vapor, small alcohol drops of 0, 2, 8 and 10 μl were injected in sequence into the container. X-ray Absorption Near-Edge Structure (XANES) spectra were taken directly after each alcohol drop, in the fluorescence detection mode using the Lytle detector. For other characterization techniques, ZnTTBPc films were spin-coated from a 12 mg/ml chloroform solution. Optical absorbance of ZnTTBPc thin films under alcohol vapor exposure were acquired in-situ by the Jenway UV-Vis spectrometer equipped with a dynamic vapor flow control. The alcohol vapor was generated by passing a nitrogen (99.9%) carrier gas into alcohol liquid at a fixed gas flow of ~750 ml/min. The Transmission Electron Microscopy (TEM) was obtained by the 120 kV Phillips CM12 microscope in both bright field and electron diffraction modes. The TEM sample preparation is described elsewhere [7]. The computational investigation [5] was performed by using the GAUSSIAN 03 DFT package with the B3LYP/6-31G* level.

Optical absorbance

a)

1.2 1.0

ZnTTBPc ZnTTBPc-MeOH

b)

0.8 0.6 0.4 0.2 0.0 400 600 800 Wavelength (nm)

Fig. 1. Optical gas-sensing properties of ZnTTBPc: a) optical absorption spectra of the spin-coated films prior to and after methanol exposure and b) changes of the absorbance in the 635–795 nm range of the thermally annealed films under dynamic alcohol flow.

Molecular Interactions Between Alcohols

3.

213

Results and Discussion

Figure 1 illustrates optical responses of the spin-coated ZnTTBPc films to various alcohols. Changes in the optical absorption spectra induced by a saturated methanol vapor are shown in Fig. 1a. However, these spectral changes were found to be irreversible. In obtaining reversible gas-sensing properties of the ZnTTBPc films, we have thermally annealed the films to 340°C to enhance their stability [5]. Figure 1b depicts optical gas-sensing responses of the annealed ZnTTBPc films to ethanol (EtOH), methanol (MeOH) and iso-propanol (IPA). The results illustrate reversible response characteristics, which are typical of a gas sensor. To obtain the gas-sensing responses, the absorption spectra were divided into wavelength regions, each of which provides one gas-sensing response signal. The response signal was calculated from the fractional change in its spectral area during the alcohol flow. In addition, the responses for the three tested alcohols differ in magnitude. This demonstrates the capability of organic optical gas sensors in distinguishing the type of alcohols [3,5]. b)

2

Energy (kcal/mol)

0 -2 -4

ZnTTBPc-IPA ZnTTBPc-EtOH ZnTTBPc-MeOH

-6 -8

ZnTTBPc-IPA ZnTTBPc-MeOH

2.2

2

3 4 5 6 7 Distance (angstrom)

Zn oxidation state

a)

2.0 1.8 1.6 1.4 1.2 1.0 0.8

-5

0 5 10 15 20 Alcohol drop size (μl)

25

Fig. 2. Weak interactions: a) DFT calculations of alcohol-ZnTTBPc interactions and b) shifts in the effective oxidation state of zinc observed by XANES measurements.

In the reversible gas-sensing regime, Fig. 2a compares the calculated interactions of ZnTTBPc with MeOH, EtOH and IPA. The net Natural Bond Orbital (NBO) charge transfers of the Zn atom induced by the alcohol interactions were +0.027e, +0.020e, and +0.014e, respectively. Figure 2b displays the XANES measurements of the zinc oxidation state obtained from shifts in the absorption edge of the spectra during alcohol exposure. Zinc metal foil and ZnSO4 were used as the oxidation number standards. The positive values of the net NBO charge transfer signify that the ability of the Zn atom to attract electrons diminishes, coinciding with the decrease in the effective oxidation state. The results demonstrate that the charge transfer mechanism may be a suitable model for the gas-sensing interactions. However, a more detailed comparison is required, particularly to account for the magnitude of the shifts. Note that an increase in the zinc oxidation state was also observed for the IPA exposure. This may be caused by other effects, such as desorbing water molecules.

214

S. Uttiya et al.

Fig. 3. Strong interactions between methanol vapor and ZnTTBPc: a) optical absorbance at 685 nm of spin-coated ZnTTBPc films during exposure to saturated methanol vapor and b) the corresponding optical phase contrast microscopic images.

In the strong interaction regime, Fig. 3a shows the optical absorbance at a fixed wavelength of 685 nm of the spin-coated ZnTTBPc films during exposure to saturated MeOH vapor. The decay of absorbance was also found to be irreversible. The optical phase contrast microscope images in Fig. 3b further explain that the absorbance changes should originate from the crystallization of ZnTTBPc induced by the saturated methanol vapor. Figure 4 displays the bright field (a) and electron diffraction (b) TEM images of ZnTTBPc films after exposure to saturated MeOH vapor. The bright field TEM image shows entangling nanometer-sized chains of ZnTTBPc. This may originate from molecular stacking, as observed for other metal phthalocyanines [8]. The sharp electron diffraction spots in Fig. 4b further illustrate the crystallization of the MeOH-annealed films. Note that no diffraction features were observed in the films as spin-coated. However, since the ZnTTBPc crystallites are randomly oriented, the electron diffraction pattern as observed originates from a few different crystal orientations. This hinders an accurate analysis to obtain the crystal structures of ZnTTBPc. Although the apparent effects of the reversible and the irreversible interaction regimes of ZnTTBPc films with alcohols are evident, the factors that initiate the latter will be need to be investigated in more details. The vapor concentration is one parameter that distinguishes the two regimes. Another important factor may be the level of pre-existing crystallization in the films. As reported in our previous work [5], the thermal annealing was found to enhance the stability of ZnTTBPc optical gas sensors, which was able to tolerate the alcohol vapor concentration of about 10 mole percents. The ZnTTBPc as spin-coated can stand much less alcohol vapor concentration. As evident in Fig. 3a, the optical absorbance of the films decays with minutes of the exposure. Nevertheless, we have also found that a prolonged exposure (>3 hr) of crystallized ZnTTBPc films to saturated alcohol vapor could led to dissolution of the films.

Molecular Interactions Between Alcohols a)

215

b)

200 nm

Fig. 4. Crystallized ZnTTBPc films after methanol annealing: a) a bright field TEM image and b) a selected area electron diffraction pattern.

4.

Conclusion

Reversible, optically active interactions between ZnTTBPc and alcohols at low vapor concentration were studied by quantum chemical calculations and x-ray absorption spectroscopy, while the irreversible crystallization of the ZnTTBPc films was observed in the films exposed to saturated alcohol vapor. Further investigation into the factors that initiate this irreversible transformation will be beneficial to improving the stability of organic gas sensors. Acknowledgements. We acknowledge the Franco-Thai project (2007), TMEC (NECTEC) and NANOTEC (Project no. NN-B-22-m1-94-49-01) under NSTDA (Thailand), and the National Synchrotron Research Center (NSRC), Thailand.

References 1 2 3 4 5 6 7 8

T.C. Pearce et al., Handbook of Machine Olfaction, Wiley-VCH, Darmstadt, 2003. M.M. Salleh, Akrajas, M. Yahaya, Thin Solid Films 417, 162, 2002. J. Spadavecchia et al., Sens. Actuators B 113, 516, 2006. G.J. Mohr, B. Bussemer, U.-W. Grummt, Sens. Actuators B 127, 414, 2007. S. Uttiya et al., J. Korean Phys. Soc., in press. W. Klysubun et al. Nucl. Instr. Meth. Phys. Res. A 582, 87, 2007. S. Pratontep et al. Synthetic Metals 146, 387, 2004. M. Brinkmann et al. Thin Solid Films 292, 192, 1997.

Organic Thin-Film Transistors with Enhanced Sensing Capabilities M. Daniela Angione1, Francesco Marinelli1, Antonio Dell’Aquila1, Alessandro Luzio2 , Bruno Pignataro3 and Luisa Torsi1*. 1

Dipartimento di Chimica – Università degli Studi di Bari – 70126 Bari – Italy * E-mail: [email protected] 2 PLAST_ICS – Superlab – Consorzio Catania Ricerche – Stradale Primosole 50 – 95121 Catania, Italy 3 Dipartimento di Chimica Fisica “F. Accascina” – Università degli Studi di Palermo – 90128 Palermo, Italy Abstract. Organic thin-film transistors, used as sensing devices, have been attracting quite a considerable interest lately as they offer advantages such as multi parameter behaviour and possibility to be quite easily molecularly tuned for the detection of specific analytes. Here, a study on the dependences of the devices responses on important parameters such as the active layer thickness and its morphology as well as on the transistor channel length is presented. To introduce the least number of variables the system chosen for this study is quite a simple and well assessed one being based on a thiophene oligomer active layer exposed to 1-butanol vapours.

1.

Introduction

Despite the impressive list of available analytical technologies, designing an inexpensive handheld or household sensor-based system is still an open challenge for the worldwide scientific community. In addition to sensitivity, selectivity and robustness, low power consumption and compact size are also stringent requirements that sensor devices have yet to fulfil [1]. The use of organic active layer such as conductive polymers (CPs) instead of metal-oxide, is being widely investigated in chemiresistors [2]. The wide range of organic materials that can be synthesized enables the fabrication of CP chemiresistors with sensitivities over a broad range of organic compounds. CPs are cost effective but reliability is still an issue. Organic thin film transistors (OTFTs) are field effect devices with organic or polymer semiconductor thin film as channel materials. They are currently one of the most fascinating technology for chemical and biological sensing both as discrete elements [3] or implemented in plastic circuits. The field of OTFTs sensors has grown in the last years thanks to the improvements achieved by different groups, according with the needs of reliable system for selective, on line chemical and biological assays. Recently such OTFT sensors have risen the interest of the scientific community for their enhanced level of performance [4]. In this configuration, highly repeatable responses were measured by properly gate-biasing the device whereas broad chemical selectivity was conferred either by covalently bound side groups [5] or by means of CP blends [6]. Besides, an important assessment of their selectivity capabilities was achieved H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4 36, © Springer Science + Business Media, LLC 2009 –

217

218

M.D. Angione et al.

with a chiral bilayer OTFT, that exhibited field-effect amplified sensitivity allowing detection of optical isomers in the tens part-per-million (ppm) concentration range [3], i.e. with a three order of magnitude sensitivity improvement, respect to the previous literature [7,8]. Besides, the on-state sensitivity enhancement is ascribable to changes occurring in the transistor channel transport and spurious effects, such as changes in contacts resistance or leakage currents, do not dominate the on-state sensor behavior [9]. In this proceedings report we focus on the investigation of some peculiar, though quite important aspects, such as the effect on the OTFT sensor response of the active layer thickness and morphology as well as of the transistor channel length.

2.

Experimental Methods

Organic Thin-Film-Transistors Fabrication and Testing. The transistor has a bottom gate device structure that consists of a highly n-doped silicon wafer coated by a SiO2 thermal oxide. The silicon substrate with a gold pad is the gate (G) contact while the silicon dioxide is the gate dielectric. The dielectric surface was covered by a α,ω-dihexyl-hexathiophene (DHα6T) thin-film deposited by thermal evaporation. Before the measurements the OTFTs were annealed under vacuum of 10–5 Torr at 100°C for 30 min. The active layer thickness ranged from 90 to 400 nm. In the deposition the organic active layer was confined into a region by shadow masking. A series of gold source (S) and drain (D) contacts were defined, by thermal evaporation through a shadow mask, directly on the organic films. A schematic of the OTFT is reported in figure 1. Transistors were produced with different channel lengths (distance between probed pads), namely: L = 0.2 mm and 1 mm) while the channel width, pads’ longer dimension (W), was kept constant at 4 mm. The sensors were fabricated and measured in a standard laboratory environment and operated at room temperature. The device was operated in the common source configuration [3]. The current flowing between two adjacent couples of source and drain pads (Ids) versus the applied source-drain voltage (Vds), at different gate biases (Vg) are the OTFT current-voltage (I-V) characteristics. On the other hand, it is named transfer characteristic, the Ids curve measured versus the applied gate voltage (Vg), at fixed The synthetic approach for preparing is reported elsewhere [9]. S

G

D

SiO2

Si

Fig. 1 DHα6T thin-film transistor sensor structure.

Sensing Measurements. The analyte, 1-butanol, in a nitrogen stream at a controlled concentration, was delivered through a nozzle, positioned over the OTFT, directly onto the active layer surface. The saturated 1-butanol vapors corresponded to a concentration of 15000 ppm. Two computer-controlled were used to achieve the

Organic Thin-Film Transistors with Enhanced Sensing Capabilities 219

waste

N2

Agilent 4155C

Fig. 2. Analyte flow control system

desired analyte concentration by diluting the saturated analyte vapors. N2 was used as inert gas during the sensors base line measurements and as carrier gas. as sketched in figure 2. Atomic Force Microscopy Measurements. DHα6T thin-films were deposited on t-SiO2/HMDS substrates also for the scanning force microscopy (SFM) inspection. SFM was carried out in air by tapping mode with a Multimode/Nanoscope IIIa. The measurements were carried out at the laboratory temperature and humidity were of about 20 °C and 40%, respectively. This system is equipped with an extended electronics module. Commercially available etched silicon probes with a pyramidal shape tip having a curvature of nominally 10 nm and an internal angle of 35° were used. The scans were performed with a the cantilever 125 µm long having a spring constant in the range of 20–100 N/m, which was oscillated at its resonance frequency (~330 kHz). Images were performed by collecting 512 × 512 points and by maintaining the scan rate below 1 Hz.

3.

Results and Discussion

Typical DHα6T OTFT I-V characteristics are reported in figure 3. The curves are relevant to OTFT of different channel length, namely 1 and 0.2 mm. The square root of Ids vs. Vg curves are reported as inset of the figures as well (figures’ 3a and 3b insets). The devices were tested as p-channel materials and the field effect mobilities in saturation regimes were extracted using the well-know equation [10]: I ds = Ci μ

W (Vg − Vt )2 2L

(1)

where Ids is the drain-source current, Ci is the capacitance per unit area of the gate dielectric layer, Vg is the gate voltage and Vt is the threshold voltage. Vt was extrapolated from the (Ids)1/2 vs. Vg plot above reported.

220

M.D. Angione et al. -100

a

0.010

1/2

A

1/2

Ids

0.004

-15 0

Ids (μA)

Ids (μA)

0.000 20

-20 -40 -60 -80 -100

Vg (V)

-40

0.003 0.002 0.001 0.000 20

0

-20 -40 -60 -80 -100

Vg (V)

-10 -5

-20 0

0.004

0.006

0.002

-60

b

0.005

-20 Ids1/2 A 1/2

0.008

-80

0

-20

-40

-60

Vds(V)

-80

-100

0

0

-20

-40

-60

-80

-100

Vds(V)

Fig. 3. DHα6T OTFT I-V characteristics and transfer characteristics for devices of different channel length: a) L = 1 mm and b) L = 0.2 mm. Relevant mobilities and threshold voltages are: a) 0.095 cm2 /Vs and –1.72 V. b) 0.11 cm2 /Vs and –5.5 V.

Channel mobilities as high as 0.08 cm2/V·s with an on/off of 2 × 103 were reached with a top contact geometry. These figures were improved by annealing the substrate film for 30 min at 100°C (vacuum ambient) resulting in a mobility ranging from 0.095 to 0.11 cm2/V·s and on/off ratio of 104. Such mobilities are extracted from the output square-roots (Ids) vs. Vds curves showed in the insets of figures 3a and 3b. Moreover leakage currents value of few nA were observed. Post-thermal annealing treatments have been known to improve molecular ordering and grain sizes of the thin film and frequently result in better device performance. Annealing can reduce also the concentration of adsorbed impurity dopants (moisture and oxygen) increasing the OTFT properties [11–13]. To quantitatively extract the contribution of the contact resistance the following experiments were performed. The I-V curves at different Vg values for two OTFTs, of different channel length (L = 0.2 and 1 mm) fabricated, contemporarily, on the same Si/SiO2 wafer, were measured for DHα6T devices in N2. The total device resistances, R, were extracted from the Ids-Vds curves in the linear region (Vds comprised between 0 and –10V) for the two OTFTs. The values for the contact resistances at different gate bias are extrapolated at L = 0, according to the transfer line method [14, 15]. The curves for the contact and total resistances exhibit contact resistance at high Vg biases falling in the low 105 Ohm × cm range. Although lower contact resistances, Rc, have been reported in some cases for other organics semiconductors such as pentacene [16, 17], values in the 104 Ohm × cm, but more often, in the 105 Ohm × cm are typical for thiophene based materials contacted with gold electrodes [17, 18]. In figure 4 the contact and channel resistances, Rch, of OTFT with different channel lengths are plotted vs.Vg. It is seen that in longer channel devices the contacts resistance weights less. In fact, the contact and channel resistances are, at worst, of comparable size. In fact, much lower contact resistance is seen, compared to channel resistances, in longer channel device. So that, contact related effects are not dominating the OTFT transport properties [14]. It has been also demonstrated that the change in channel resistance is also much higher than that of the contacts, proving that the sensing effect occurs in the channel region and not at the contacts [9].

Organic Thin-Film Transistors with Enhanced Sensing Capabilities 221

R (MΩ )

100

10 Rc Rch (L = 0.2 mm) Rch (L = 1 mm)

1 -20

-40

-60

Vg (V)

-80

-100

Fig. 4. DHα6T contact and channel resistances as extracted from the relevant I-V curves.

Figure 5 reports on SFM images showing the topography of DHα6T thin films at different thickness. The 90 nm sample shows a network of laterally interconnected planes with the presence of screw dislocations. The screw base typically is of the order of hundreds of nanometers. The planes are mainly parallel to the dielectric substrate and show step heights of about 4.5 nm, which is consistent with the simulated full length of a DHα6T. These films exhibit surface square mean roughness (RMS) of 3.2 nm. In the 200 nm thick sample the growth of screw dislocations is observed too. The planes are randomly oriented in space and appear to be quite interconnected, i.e. at this thickness a lateral continuity of the whole system is preserved. The RMS roughness of this sample is only slightly changed and the measured step heights closely resembles that of the 90 nm thin sample. Finally, in the thicker sample (400 nm) the screw-mode growth leads to tip-structures about 60–80 nm large along with a loss of interconnection is observed. The step heights are still of molecular size, whereas the RMS roughness increases to 5.2 nm. The above findings are consistent with a picture in which the 3D growth characterized by screw dislocations allows for an increase of the active-surface area for the gas physi/chemi-sorption. According to this scenario, thicker films are characterized by a significant surface area gradient along the bulk depth profile which in turn should lead to an absorption gradient along with an analyte concentration loss ongoing towards the two dimensional transport layer. An increased interconnection of the 2D layer at the interface with the gate dielectric, (where the field induced transport is known to take place) is indeed expected for thicker films, due to voids filling by incoming material during the deposition process. A compromise between film thickness and sensing response is thus expected. The evaluation of the sensing properties was achieved by measuring Ids-Vg transfer characteristics in N2 and in a flux of saturated vapors of 1-butanol. A typical example is reported in figure 6a. Figures 6b and 6c show standard transistor characteristics measured while the device was exposed to N2 or to 1-butanol (11250 ppm). Generally, as it is apparent in figure 6 and as already widely reported, a onstate Ids decrease occurs upon exposure of a thiophene based OTFT to alcohol vapors [19].

222

M.D. Angione et al.

Fig. 5. SFM images of DHa-6T films at thickness of: a) 90 nm, b) 200 nm, c) 400 nm. At the bottom zoom typical features of the films are reported. Note the screw-mode growth leading to tip-structures for the thicker samples.

Fig. 6. DHα6T OTFT I-V transfer characteristics (a) and characteristics for a L=0.2 mm channel length measured in N2 (b) and in 1-butanol atmosphere (c).

This can be explained by considering the potential barrier increase at grain boundaries upon exposure to the analyte [2]. As a matter of fact, upon exposure of the OTFT active layer to a volatile organic vapor a partition of the analyte molecules between the solid and the gaseous phase occurs. The chemical affinity between the analyte and the active layer modulates the degree of physi/chemisorption of the analyte molecules at the gain surface. This in turns, enhances the potential barriers at the boundaries, eventually lowering the intensity of the drifting source-drain current, as seen in preliminary evidences [20]. This effect involves the whole bulk of the DHα6T film, therefore also the interface with the gate dielectric, eventually affecting the two-dimensional transport occurring in this region [21].To analyze the data the response of the OTFT sensor ΔIds was defined as follows: ΔI ds = ( I ds ( N 2 ) − I ds (1 − bu tan ol ) )

(2)

Organic Thin-Film Transistors with Enhanced Sensing Capabilities 223

clearly a different Δ I ds value can be measured at different Vg biases. As a general trend ΔIds increases by several order of magnitudes while the devices is driven from the off state (Vg
Fig. 7. Ids-Vg transfer characteristics of DHα6T OTFT exposed to N2 (continuos line) and butanol (dotted line) by using active layer of thickness of 90 nm (a), 200 nm (b) and 400 nm (c).

This is confirmed also by the slightly lower mobility. In the thicker film a diffusion limited effect could explain the data. In other words, the observed films structure is particularly consistent with a picture in which the thicker is the film the larger is the structural interconnection, but the lower is the analyte amounts, reaching the two dimensional transport channel.

4.

Conclusions

Herein the electrical characteristics and sensing properties of DHα6T OTFT have been investigated. The devices have shown field effect mobility of 0.11 cm2/V·s with low leakage currents and fair contact resistance. DHα6T films have been exposed to saturated vapours of 1-butanol revealing good performances when employed as gas sensor. Indeed, the sensing response was found to be thickness/ morphology dependent when films in the 90–400 nm thickness range were investigated. The sensor response measured at a controlled concentration of 1-butanol was higher for the 200 nm thick film whereas a lower sensor response was observed for the thicker (400 nm) and thinner films (90 nm). Although the 400 nm thick one resulted to have the highest sensing active area (AFM), its thickness reduced the analyte diffusion to the channel active area eventually reducing the electric response.

224

M.D. Angione et al.

Acknowledgements. This work was partially supported by the “PRIN-06 Project 2006037708 - Plastic bio-FET sensors”.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

K.C. Persaud, Mater. Today, 8, 38–44, 2005. L. Torsi, A. Dodabalapur, Anal. Chem., 77, 380A–387A, 2005. L. Torsi, G.M. Farinola, F. Marinelli, M.C. Tanese, O.H. Omar, L.Valli, F. Babudri, F. Palmisano, P.G. Zambonin, F. Naso Nat. Mater., 7, 412–417, 2008. K.C. See, A. Becknell, J. Miragliotta, H.E. Katz, Adv. Mater., 19, 3322, 2008. L. Torsi. et al., J. Phys. Chem. B, 107, 7589, 2003. J. Huang, J. Miragliotta, A. Becknell, H.E. Katz, J. Am. Chem. Soc., 129, 9366, 2007. E. J. Severin et al., Anal. Chem., 70, 1440, 1998. B.P.J. de Lacy Costello, N.M. Ratcliffe, P.S. Sivanand, Synth. Met., 139, 43, 2003. L. Torsi, F. Marinelli, M.D. Angione, A. Dell’Aquila, N. Cioffi, E. De Giglio and L. Sabbatini, JACS, submitted. G. Horowitz, Adv. Mater., 10, 365-376, 1998. M. Stolka, M.A. Abkowitz, Synth. Met., 54, 417, 1993. H. Sirringhaus, N. Tessler, R. H. Friend, Science, 280, 1741-1744, 1998. L. Torsi, A. Dodabalapur, A.J. Lovinger, H.E. Katz, R. Ruel, D.D. Davis, K.W. Baldwin, Chem. Mater., 7, 2247, 1995. Panzer, Friesbie C.D., in Organic Field-Effect Transistors (Eds: Z. Bao, J. Locklin) CRC Press Taylor & Francis Group, Ch. 2.4, 2007. J. Zaumseil, K.W. Baldwin and J.A. Rogers, J. Appl. Phys., 93, 6117, 2003. P.V. Pesavento, K.P. Puntambekar, C.D. Frisbie, J.C. McKeen, P.P. Ruden, J. Appl. Phys., 99, 094504, 2006. E.J. Meijer, G.H. Gelinck, E. van Veenendaal, B.-H. Huisman and D.M. de Leeuw, T.M. Klapwijk, Appl. Phys. Lett., 82, 4576–4578, 2003. B.H. Hamadami and D. Natelson, J. Appl. Phys., 97, 064508, 2005. T. Someya et al., Appl. Phys. Lett., 81, 3079–3081, 2002. A. Dodabalapur, SPIE Optics and Photonics – August, San Diego, CA, 2007. A. Dodabalapur, L. Torsi, H.E. Katz, Science, 268, 270-271, 1995.

Photoelectric Properties of Microrelief Organic/Inorganic Semiconductor Heterojunctions N.L. Dmitruk1, O.Yu. Borkovskaya1, D.O. Naumenko1, I.B. Mamontova1, N.V. Kotova1, O.S. Lytvyn1, and Ya.I. Vertsimakha2 1

Institute for Physics of Semiconductors, NAS of Ukraine, 45 Nauki Prospect, Kyiv 03028, Ukraine E-mail: [email protected] 2 Institute of Physics, NAS of Ukraine, 46 Nauki Prospect, Kyiv 03028, Ukraine E-mail: [email protected] Abstract. The effect of texturing interface on optical and photoelectric properties of organic/inorganic semiconductor heterostructures, based on n-GaAs, has been investigated both for anisotype (with pentacene or lead phthalocyanine) and isotype (with methyl perylene pigment) junctions. The analysis of internal quantum efficiency spectra, calculated from the experimental short-circuit photocurrent and light reflectance spectra, allowed to determine the participation of the OS layer in the formation of heterostructure total photocurrent. Considerable enhancement of the photocurrents, generated in both organic and inorganic semiconductors was found for all the investigated heterojunctions.

1.

Introduction

Organic semiconductors (OS), such as pentacene, lead phthalocyanine etc. have attracted considerable attention as the components of barrier heterojunctions with other organic or inorganic semiconductors (IS) for transistor, sensor or photovoltaic application. Therefore, dependence of the electronic and recombination properties of the heterostructure interface on the technological conditions of the interface formation is of great importance [1–6]. This work is devoted to investigation of the effect of texturing organic/ inorganic semiconductor interfaces on the optical, photoelectric and electrical properties of barrier OS/GaAs heterostructures in search of the more efficient method of the OS/IS interfaces modification for the enhancement of the heterostructure photosensitivity. Both anisotype and isotype junctions, based on n-GaAs, were fabricated with pentacene (Pn) and lead phthalocyanine (PbPc) as organic semiconductor of p-type, and methyl perylene pigment (MPP) as OS of n-type.

2.

Experimental Methods

OS/IS heterostructures were fabricated by the OS layers evaporation in vacuum (2·10–4 Pa) at room temperature (~300 K) on preliminary prepared flat and textured

H.-G. Rubahn et al., Interface Controlled Organic Thin Films, DOI: 10.1007/978-3-540-95930-4–37, © Springer Science + Business Media, LLC 2009

225

226

N.L. Dmitruk et al.

(100) substrates of n-GaAs. The GaAs surface treatment included chemical polishing or anisotropic etching (in 2HF:2H2SO4:1H2O2 solution) to obtain flat or textured (with microrelief of quasigrating type) interface. Photodiode structures were fabricated by evaporation of semitransparent Au layer onto OS film and of In onto n-GaAs back surface to obtain ohmic contact. For the comparison Au/IS Schottky diodes were fabricated simultaneously on the same substrates. The surface microrelief morphology (of polycrystalline type) of deposited films was investigated by AFM technique. The reflectance spectra at a number of angles of incidence for p- and s-polarized light allowed calculating the absorption coefficient spectral dependencies and the spectra of light transmittance in layered barrier structures, photoelectric properties of which have been investigated in the spectral range from 400 to 950 nm.

3.

Results and Discussion

flat

0,4

rel

3 0,2

2 0,0 400

8

a

1

b

2

6

1,0

5

4

4 3

0,5

1

2 600

800

Wavelength, nm

1,5

400

α, a. u.

0,6

Iph / Iph

External quantum eff.

The spectra of the short-circuit photocurrent, expressed as quantum efficiency, for anisotype (and isotype) OS/n-GaAs heterostructures with a different spectral range of the light absorption by OS layer are shown in Fig. 1a (and Fig. 2a). Corresponding spectra of the OS/IS heterostructure photocurrent enhancement due to texturing interface are shown in Figs. 1b and 2b. together with the spectra of OS layer absorptance (α).

600

800

0,0

Wavelength, nm

Fig. 1. The external quantum efficiency spectra (a) and spectra of the iphrel/iphflat ratio (b) for Au/GaAs (1), Au/PbPc/GaAs (2) and Au/Pn/GaAs (3) heterostructures, and the absorptance for PbPc (4) and Pn (5). The layer thickness, nm: 18 (Au), 51 (PbPc), 79 (Pn).

It is seen, that the PbPc/n-GaAs heterostructure have the least photosensitivity both for flat, and textured interface in spite of the greatest increasing of photocurrent (iph) due to texturing interface. The spectral dependences of iphrel/iphflat and α for this structure are antibatic. So, the value of its photocurrent is mainly determined by photogeneration in GaAs layer weakened by the absorption of light in PbPc layer. The increase of the short-circuit photocurrent for Pn(MPP)/n-GaAs structures due to texturing interfaces is much greater than for Au/n-GaAs structure. This can be caused by an additional increase of the light absorption in OS layer due to multiple internal reflection of light from textured interface, and on the contrary, by

a

0,00 400

4 2

600

800

Wavelength, nm

0 400

227

5

1,5x10

2 -1

6

α, cm

1

0,02

flat

0,04

b

8

rel

0,06

Iph / Iph

External quantum eff.

Photoelectric Properties of Microrelief Organic

1

5

1,0x10

4

5,0x10

600

800

0,0

Wavelength, nm

Fig. 2. The external quantum efficiency spectra (a) and spectra of the iphrel/iphflat ratio (b) for Au(36 nm)/MPP(80 nm)/GaAs (1), and the absorptance for MPP (2).

increase of the light transmittance into n-GaAs, in particular, in the spectral range, where absorption in OS layer is absent. The considerable decrease of the series resistance also furthers to the enhancement of the photocurrent.

4.

Conclusions

The comparative investigation of photoelectric and optical properties of OS/nGaAs heterostructures (p-Pn, p-PbPc, n-MPP/n-GaAs) and the analysis of their internal quantum efficiencies allowed to distinguish the contribution of OS layer to the generation of total photocurrent and to determine the most photosensitive heterostructures. Photosensitivity and efficiency of isotype MPP/n-GaAs heterostructures was shown to be greater than ones of anisotype Pn and PbPc/n-GaAs structures. Texturing of OS/IS heterostructure interface by wet chemical anisotropic etching of IS substrate resulted in a considerable increase of the photocurrent for all the investigated structures. So, the developed texturing of OS/IS heterostructures interfaces may be perspective for photovoltaic application after the optimization of their parameters.

References 1 2 3 4 5 6

Ya.I. Vertsimakha, Mol. Cryst. Liq. Cryst., 355, 275, 2001. Ya. Vertsimakha, P. Lutsyk, Mol. Cryst. Liq. Cryst., 467, 107, 2007. D. Faltermeier, B. Gomof, M. Dressel, A.K. Tripathi, J. Pflaum, Phys. Rev. B., 74 125416, 2006. N.L. Dmitruk, I.B. Mamontova, O.Yu. Borkovskaya, Ya.I. Vertsimakha, Mol. Cryst. Liq. Cryst., 384, 49, 2002. S.J. Kang, Y. Yi, C.Y. Kim, S.W. Cho, M. Noh, K. Jeong, C.N. Whang, Synthetic Metals, 156, 32, 2006. Ya. Vertsimakha, P. Lutsyk, K. Palewska, J. Sworakowski, O. Lytvyn. Thin Solid Films, 515, 7950, 2007.