The influence of structure on the electronic properties of $\pi$-conjugated systems

Cell-phones, computers, displays and lighting are an important part of everyday life. To increase the efficiency, functionality and to reduce production costs of these devices, the development of new functional electronic materials is essential. Carbon-based materials have the potential to provide many of the needed functionalities. That is, organic chemistry is able to provide a large number of different molecules consisting of primarily carbon which offer tunable electronic properties. This way, molecules can be used to make organic light emitting diodes (OLEDs) for lighting and displays, organic solar cells for energy harvesting and even micro electrical mechanical systems (MEMS) that tell your smart phone which side is up. Carbon can also be grown in single layers called graphene. Graphene offers the potential of high-speed logic and may be the basis of completely new types of switching devices such as (pseudo)-spintronic devices. The key to creating functional electronic materials from carbon is getting the right structure. Therefore this thesis focusses on the influence of the structure on the electronic properties of p-conjugated systems (i.e. graphene, electronic functionalized molecules, etc.). In the first part of this thesis this influence is studied for a graphene layer grown on silicon carbide (SiC), in which the graphene/SiC interface has a strong influence on the electronic properties, and for the formation of a graphene layer by decomposing C60 on a Pt(111) surface, where the exact process of decomposition results in a particular graphene/Pt(111) interface structure. The second part of this thesis deals with the polymer system poly(3,4-ethylenediox-thiophene):poly(styrenesulfonate) (PEDOT:PSS). PEDOT: PSS is a transparent conducting material which can be used in OLEDs and organic solar cells, where it is typically applied by spin coating from emulsion. Here PEDOT and PSS are found to form filaments, resulting in a spaghetti-like morphology of the spin cast layer, which has important consequences for the charge transport. In Chapter 2, graphene grown by thermal decomposition of SiC is studied by atomically resolved scanning tunneling microscopy and spectroscopy (STM/STS). At small bias voltages STM images reveal the graphene lattice structure, as expected. However, at increased bias voltages bright features are revealed which, by comparing STS to angular resolved photo emission spectroscopy (ARPES), are shown to be localized states of the interface layer between the graphene and the SiC. Additionally, close to structural defects in the graphene layer a giant inelastic tunneling process, caused by electron-phonon coupling, is observed. This process accounts for half the total tunneling current. A map of this inelastic current shows that these inelastic contributions are strongest at the localized states of the interface layer. Therefore it is expected that the localized states of the interface layer have an important yet complicated influence on the electronic properties of graphene on SiC. In Chapter 3, the full pathway from room temperature deposition of C60 on Pt(111) to the formation of graphene at high temperatures is presented. Using in-situ low temperature STM, a submonolayer of C60 on Pt(111) is studied after heating steps at increasing temperatures. We are able to identify the molecular orientations of the C60 molecules at each step. Changes in the apparent height of C60 molecules in combination with a change in their orientation show that C60 molecules adopt a subsurface missing-atom configuration. By studying the graphene layer formed upon an additional high temperature heating step, both by STM and by density functional theory (DFT) calculations, we show that v3xv3R30° domains of the formed graphene layer are also in a subsurface missing-atom configuration. It is shown that the transition toward the subsurface missing-atom configuration takes place at the edges of C60 islands and could be assisted by the instability of the Pt(111) surface. In Chapter 4, the mechanism and magnitude of the in-plane conductivity of PEDOT:PSS thin films is determined using temperature dependent conductivity measurements for various PEDOT:PSS weight ratios. For all studied weight ratios the conductivity of PEDOT:PSS is well described by quasi 1D variable range hopping (VRH). The experimentally determined conductivity varies over three orders of magnitudes and follows a power law with power 3.5 as a function of the weight fraction of PEDOT in PEDOT:PSS in the range 0.04-0.3. Analysis of the field dependent conductivity shows a behavior that is consistent with quasi-1D VRH. Combined, these observations suggest that conductance takes place via a percolating network of quasi-1D filaments. Using transmission electron microscopy (TEM) filamentary structures are indeed observed both in vitrified solutions and in dried films. For PEDOT:PSS films that were processed with a high boiling solvent, the temperature dependence of the Ohmic conductivity suggests a quasi-1D VRH system, but the low characteristic temperature indicates that the system is close the critical regime between a metal and an insulator. In this case, the conductivity scales linearly with the weight fraction of PEDOT in PEDOT:PSS, indicating the conduction is no longer limited by a percolation of filaments. The lack of observable changes in TEM upon addition of the high boiling solvent suggests that the changes in conductivity are due to a smaller spread in the conductivities of individual filaments, or a higher probability for neighboring filaments to be connected, rather than being due to morphological modification of the filaments. In Chapter 5, the room temperature out-of-plane conductivity of spin coated PEDOT:PSS films is studied. Although important for its application as transparent conductor in light emitting and photovoltaic devices, studies to the conductivity of PEDOT:PSS rarely address the out-of-plane conductivity and those that do report widely varying results. In the presented experiment, the out-of-plane charge transport in thin films of PEDOT:PSS is systematically studied by varying its composition. To this end, small vias between metallic contacts are used. An unexpected, but strong dependence of the conductivity on via diameter is observed. The change in conductivity correlates with a diameter dependent change in PEDOT:PSS layer thickness. The more than three orders of magnitude variation in out-of-plane conductivity with only a 3-4-fold layer thickness variation can quantitatively be explained on basis of a percolating cluster model. This model describes the probability for conductive paths between the top and bottom electrode to be formed from randomly placed conductive elements in an insulating matrix, and shows that for thin layers this probability strongly decreases with increasing layer thickness. The results also rationalize previously unexplained findings in molecular junctions where PEDOT:PSS is used as contact electrode. In Chapter 6, the out-of-plane conductivity of spin coated PEDOT:PSS films is studied for varying temperatures, electric fields and composition. Recent measurements of the out-of-plane conductivity of PEDOT:PSS, and of the conductivity of a semiconducting polymer in a field effect device reveal a curious power law dependence of the conductivity on field and temperature in both the Ohmic and non-Ohmic regimes. This is referred to as universal scaling. Connecting this behavior to a particular microscopic model has proven to be difficult. Here we have performed a systematic study of the temperature and bias voltage dependence of the out-of-plane conductivity of PEDOT:PSS through the use of interconnect structures (vias) for varying via diameters and various PEDOT:PSS formulations. The measurements indeed show universal scaling. By use of explicit knowledge of the microscopic structure of the used PEDOT:PSS materials the number of possible underlying models can be narrowed down to only three models: a model for finite size effects in quasi one-dimensional variable range hopping, a model for a chain of quantum dots in the Coulomb blockade regime and a model for connected Luttinger liquids. The presented measurements seem at odds with all but the latter model. Summarizing, this thesis characterizes the relationship between the electronic properties of p-conjugated systems and their structural properties for two systems, graphene and PEDOT:PSS. For graphene, this is shown for two individual cases, one discusses the influence of localized states of an interface layer below graphene on SiC, the other discusses the formation of a missing atom interface structure of the Pt(111) surface below a graphene layer. The second part of this thesis presents a concerted effort studying the charge transport properties of PEDOT:PSS. Starting from its structure, which is found to consist of a random (percolating) network of filaments, a very complete set of measurements could be accurately described. This includes the in- and out-of-plane conductivity of PEDOT:PSS as a function of temperature, electric field and PEDOT:PSS ratio. The uncovered filamentary structure differs from the current literature which typically describes PEDOT:PSS as consisting of grains with a PEDOT-rich core and PSS-rich shell. Given the lower percolation threshold for filaments than for grains, filamentary PEDOT:PSS is likely to have a higher conductivity than grain-like PEDOT:PSS. It is therefore not unlikely that with the efforts to increase the conductivity of PEDOT:PSS, changes have been made in the synthesis of commercial PEDOT:PSS to produce filamentary instead of grain-like PEDOT:PSS. The presented model for the conductivity of PEDOT:PSS offers a good starting point for further rational optimization of the properties of PEDOT:PSS