Efficient characterization of bulk heterojunction films by mapping gradients by reversible contact with liquid metal top electrodes

The ways in which organic solar cells (OSCs) are measured and characterized are inefficient: many substrates must be coated with expensive or otherwise precious materials to test the effects of a single variable in processing. This serial, sample-by-sample approach also takes significant amounts of time on the part of the researcher. Combinatorial approaches to research OSCs generally do not permit microstructural characterization on the actual films from which photovoltaic measurements were made, or they require specialized equipment that is not widely available. This paper describes the formation of one- and two-dimensional gradients in morphology and thickness. Gradients in morphology are formed using gradient annealing, and gradients in thickness are formed using asymmetric spin coating. Use of a liquid metal top electrode, eutectic gallium− indium (EGaIn), allows reversible contact with the organic semiconductor film. Reversibility of contact permits subsequent characterization of the specific areas of the semiconductor film from which the photovoltaic parameters are obtained. Microstructural data from UV−vis experiments extracted using the weakly interacting H-aggregate model, along with atomic force microscopy, are correlated to the photovoltaic performance. The technique is used first on the model bulk heterojunction system comprising regioregular poly(3-hexylthiophene) (P3HT) and the soluble fullerene derivative [6,6]-phenyl C₆₁ butyric acid methyl ester (PCBM). To demonstrate that the process can be used to optimize the thickness and annealing temperature using only small (≤10 mg) amounts of polymer, the technique was then applied to a bulk heterojunction blend comprising a difficult-to-obtain low-bandgap polymer. The combination of the use of gradients and a nondamaging top electrode allows for significant reduction in the amount of materials and time required to understand the effects of processing parameters and morphology on the performance of OSCs.