30 Science Research Writing semiconductors from volatile organic precursors [11]. Here, a high vapour pressure compound (typically a metal halide or a metallorganic) of each respective metal is carried independently, via a carrier gas, to a high temperature reaction zone. In this zone, the compounds are deposited onto a heated substrate where they thermally decompose and react to yield the desired III-V compound. Th e excess reactants and reaction products are then exhausted from the system via a scrubber. In this paper we apply the techniques of VPE to grow fi lms of DAST by the reaction of two volatile organic materials in a hot-wall, atmospheric pressure reactor. By nuclear magnetic resonance (NMR) analysis, we fi nd that the stoichiometry of polycrystalline DAST fi lms is >95% pure (limited by instrumental sensitivity). Using X-ray diff raction and other analytical techniques, we observe a signifi cant dependence of fi lm quality, such as ordering and crystallite size, on the substrate composition and other deposition conditions used for growth, suggesting that it may be possible to generate optical quality thin fi lms of DAST and similar organic salts and compounds by OVPD using suitable substrates. To our knowledge, this is the fi rst demonstration of the deposition of ordered thin fi lms of a highly non-linear optically active organic salt using atmospheric vapour phase techniques. Limitations of charge-transfer models for mixed-conducting oxygen electrodes INTRODUCTION Traditionally, electrochemistry is concerned with charge-transfer reactions occurring across a 2-dimensional interface. Indeed, at any macroscopic two-phase boundary, the magnitude, direction and driving force for current density can be described relatively unambiguously. As early as 1933 [1], workers began introducing the concept of a 'three-phase boundary' (solid/liquid/gas) in order to allow for direct involvement of gas-phase species at an electrochemical interface. However, since matter cannot pass
Introduction — Writing Task 31 through a truly one-dimensional interface among three phases, concepts of 'interfacial area', 'current density', and 'overpotential' at a three-phase boundary lack clear defi nition. For example, where exactly is the current fl owing from/to, and what is the local fl ux density? Also, if we defi ne overpotential in terms of thermodynamic potentials of species outside the interfacial region, what species and region are we talking about? Although the three- phase boundary concept may serve as a useful abstraction of the overall electrode reaction, it does not address these mechanistic questions. Workers studying gas-diff usion electrodes in the mid- 1960s recognized the limitations of the three-phase boundary concept [2, 3]. As an alternative, they began to break down the electrode reaction into individual steps, some that involve charge- transfer across a two-dimensional interface, and some that involve dissolution and diff usion of molecular species in three dimensions or across a chemical interface. Th ese and subsequent studies have demonstrated that electrodes with i-V characteristics indicative of charge-transfer limitations (eg. Tafel behaviour) can, in fact, be limited by steps that do not themselves involve charge- transfer [4]. Although the solid-state literature has held on to the three-phase boundary concept more tightly than the aqueous or polymer literature, few examples remain today or solid-state electrochemical reactions that are not partially limited by solid- state reaction and diff usion processes. One example is the O2-reduction reaction on a mixed- conducting perovskite electrode, which defi es rational explanation in terms of interfacial impedance. In order to incorporate non- charge-transfer eff ects, workers oft en apply an empirical Butler– Volmer model (for DC characteristics) or an equivalent-circuit model (for AC impedance) that treat non-charge-transfer processes in terms of an eff ective overpotential/current relationship [5, 6]. However, this approach lacks generality and can oft en be incorrect for treating oxygen absorption and solid-state and gaseous diff usion, which contribute to the impedance in a convoluted manner [7]. Although such models may provide a useful set of parameters to 'fi t' data accurately, they leave the electrode reaction