The analysis of polymer interfaces : a combined approach

Abstract

This chapter describes the applicability of the surface science technique secondary ion mass spectrometry (SIMS) on polymer/organic surfaces without going into great detail. Mostly, some problems that can be encountered in applied surface science and that are described in this thesis will be discussed to provide a framework for the rest of the thesis. 1.1 Setting the stage Although bulk properties determine the strength and durability of materials, the surface properties are often of greater importance for many sophisticated applications. The material at the outermost layer may be oxidized or contaminated with organic or inorganic compounds, which can have catastrophic results for the performance in the desired application. This is certainly the case for organic solar cells and light emitting devices, in which electron transfer across interfaces may be inhibited, or polymer degradation may occur. Close to the surface, material may have diffused into the polymer structure, which can have either a negative, or sometimes a positive influence on the device performance. Either way, it will be important for the scientist to know the chemical composition in the surface region in order to improve his understanding of the system. An analytical technique that is capable of analyzing materials from the outermost atomic layer up to a depth of several micrometers is secondary ion mass spectrometry (SIMS), which will be discussed in detail in chapter 2. With SIMS, interface reactions, diffusion and segregation processes can be studied at an often remarkable depth resolution of only a few nanometers. However, this technique still has many limitations, especially for studying polymer and organic materials, which will be discussed in detail in later chapScope of this thesis 2 ters. This chapter will only set the stage by discussing the background of the studies presented in this thesis, and the objectives of each study. 1.2 Surface modified coatings Although many polymers have excellent bulk properties, their use in many applications is inhibited by their undesirable surface properties, such as poor wettability and abrasion or chemical resistance. Fortunately, there are many ways to modify the polymer surface so that it exhibits more preferable surface properties. The range in which this modification extends within the polymer layer may be limited to only a few nanometers, depending on the applied modification technique. Although this depth scale is eminently suitable for application of SIMS in combination with depth profiling, in this range the SIMS signals are not very stable and any conclusions drawn from measurements within this unstable range have to be scrutinized with suspicion. Chapter 3 describes a technique with which this instability can be avoided in the range of interest. 1.3 Polymeric light emitting displays Some polymers, such as polyacetylene (PA) that was the first such polymer reported, can act as a semiconductor because it is p-conjugated. This means that the molecular backbone, in the case of PA the whole polymer, consists of alternating single and double bonds. The p-electrons can be easily excited into delocalised molecular orbitals along the polymer chain, which causes semiconducting behavior. After PA, many other p-conjugated polymers have been reported to act as a semiconductor. One of these polymers, poly(phenylenevinylene) or PPV, was used to construct the first polymer light emitting display. Fig. 1 shows a common design for such a device. It consists of the active layer, sandwiched between two electrodes, one of which obviously has to be transparent. The chemistry of the transparent electrode often damages the active layer, so that most often a protective conductive polymer is placed between the transparent electrode and the active polymer, also to prevent the transparent electrode to make direct contact with the opposite electrode, because the transparent electrode usually is very rough and the active polymer layer rather thin. By applying a sufficiently high bias voltage, holes and electrons are injected into the polymer from both electrodes. Because these holes and electrons move towards each other under the influence of the electrical field, they may recombine on a p-conjugated polymer chain to form excitons.Subsequent radiative decay of the excitons results in the emission of light. Polymeric light emitting displays 3 The color of this light depends on the energy gap, also called band gap, between the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) of the p-conjugated polymer. This means that the color can be tuned by making some changes to the chemical structure of the polymer. For example, PPV emits yellow-green light whereas alkoxy-substituted PPV derivates emit red light. For optimal hole and electron injection the right electrode material has to be selected. For the hole injection electrode usually indium tin oxide (ITO) is used, because of its transparency and its availability from the liquid crystal display technology. However, ITO easily degrades PPV and its derivatives. To prevent this from happening, nowadays a conducting polymer layer, made of polystyrenesulfonate-doped polyethylene dioxythiophene (PEDOT:PSS) is added between the active and ITO layer, which increases the device lifetime by a factor of at least ten. To study the effectiveness of this approach, the diffusion of indium can be studied by measuring the relative concentration profile of indium with SIMS. The diffusion of aluminum, that is sometimes used as low workfunction electrode, into the active layer, a process that some suspect happens during the thermal evaporation on top of the active layer, can also be studied. However, the depth resolution obtained in metal-polymer stacks such as organic light emitting displays can be extremely bad. Chapter 4 discusses some meaFigure 1. Common structure of an organic light emitting device: a transparent electrode on a glass substrate, covered with a conducting polymer layer, on top of which the active layer has been deposited, covered with a low workfunction metal electrode. Note that the conducting polymer layer prevents the rough transparent electrode to make direct contact with the opposite electrode Scope of this thesis 4 sures that can be taken to prevent some loss of depth resolution from happening, and a technique to improve the depth resolution post-hoc. Many low workfunction metals have been tried as the electron injection electrode, such as Al, Mg and Ca. Their low workfunction ensures a good balance between the hole and electron injection, which in turn results in a higher number of photons emitted per injected charge carrier. However, these metals are susceptible to oxidation, especially at the interface, which can degrade their contact with the active polymer layer. The problem is to reach the interface with an analytical tool without perturbing the interface. However, by simulating the interface with a sub-nanometer layer of Ca on top of the polymer the oxidation mechanism at the interface could be studied and is reported in chapter 5 of this thesis. The performance of these devices can be improved remarkably by the insertion of a 1 nm thin layer of LiF sandwiched between the low workfunction metal electrode and the active polymer layer. The precise mechanism of this improvement has been the subject of considerable debate, with two mechanisms being predominantly favored. By simulating the interface with subnanometer layers of LiF and Al, one of these mechanisms could be studied and disproved, which is described in detail in chapter 6 of this thesis. 1.4 Organic solar cell panels In organic solar cells the reverse process of the polymer light emitting displays is used, so that the same basic design from the organic light emitting devices can be used. Light that falls on the solar cell results in an exciton, which relaxes through the production of a hole and electron. However, if the electron and hole are not quickly separated, they will recombine and no useful electricity will be produced by the solar cell. A molecule that can quickly adsorb and transport the electron has to be mixed with the active layer. Fullerenes are suitable as an electron acceptor, because the electron transfer from the active layer, the electron donor, has been measured to occur within fifty femtoseconds, while the reverse process takes milliseconds. The contact surface between active polymer and the electron-accepting molecule has to be as large as possible to facilitate the electron transfer. However, fullerenes have the tendency to cristallise when blended with the electron donating polymer, for example PPV. The cristallisation of the fullerenes in the blend lowers the contact surface between the donor and acceptor, and hence is undesirable. To prevent cristallisation a fullerene derivative (PCBM) was developed with a reduced tendency to cristallise, and that is somewhat easier to dissolve in solvents used for spincasting. However, PCBM may still segregate into clusters within the polymer layer that adsorb preferentially at the surface of the polymer. Organic solar cell panels 5 A depth profile of a PCBM:MDMO-PPV blend could possibly shed some light on this issue. However, the problem with applying SIMS on a blend of two organic molecules is that the raw data obtained with this technique is not particularly useful to distinguish between both molecules. Chapter 7 describes how two multivariate statistical analysis techniques, canonical discriminant analysis and principal components analysis, can be applied to distinguish between both molecules and even quantify their distribution in-depth, be it with some error. Scope of this thesis