RESUMO
Electron beam lithography (EBL) is a highly precise, serial method for patterning surfaces. Positive tone EBL resists enable patterned exposure of the underlying surface, which can be subsequently functionalized for the application of interest. In the case of widely used native oxide-capped silicon surfaces, coupling an activated silane with electron beam lithography would enable nanoscale chemical patterning of the exposed regions. Aminoalkoxysilanes are extremely useful due to their reactive amino functionality but have seen little attention for nanopatterning silicon surfaces with an EBL resist due to background contamination. In this work, we investigated three commercial positive tone EBL resists, PMMA (950k and 495k) and ZEP520A (57k), as templates for vapor-phase patterning of two commonly used aminoalkoxysilanes, 3-aminopropyltrimethoxysilane (APTMS) and 3-aminopropyldiisopropylethoxysilane (APDIPES). The PMMA resists were susceptible to significant background reaction within unpatterned areas, a problem that was particularly acute with APTMS. On the other hand, with both APTMS and APDIPES exposure, unpatterned regions of silicon covered by the ZEP520A resist emerged pristine, as shown both with SEM images of the surfaces of the underlying silicon and through the lack of electrostatically driven binding of negatively charged gold nanoparticles. The ZEP520A resist allowed for the highly selective deposition of these alkoxyaminosilanes in the exposed areas, leaving the unpatterned areas clean, a claim also supported by contact angle measurements with four probe liquids and X-ray photoelectron spectroscopy (XPS). We investigated the mechanistic reasons for the stark contrast between the PMMA resists and ZEP520A, and it was found that the efficacy of resist removal appeared to be the critical factor in reducing the background functionalization. Differences in the molecular weight of the PMMA resists and the resulting influence on APTMS diffusion through the resist films are unlikely to have a significant impact. Area-selective nanopatterning of 15 nm gold nanoparticles using the ZEP520A resist was demonstrated, with no observable background conjugation noted in the unexposed areas on the silicon surface by SEM.
RESUMO
Organic photovoltaics (OPVs) are assembled from a complex ensemble of layers of disparate materials, each playing a distinct role within the device. In this work, the role of the interface that bridges the transparent anode and the bulk heterojunction (BHJ) in an OPV device was investigated. The surface characteristics of the electrode interface affect the energy level alignment, phase segregation, and the local composition of the bulk heterojunction (BHJ), which is in close contact. The commonly used ITO/PEDOT:PSS electrode was tailored with a thin, low-band-gap polymer overlayer, called PBDTTPD-COOH, a variant of the established donor polymer, PBDTTPD. Three BHJs that were composed of a donor polymer and PC71BM, were examined, including the donor polymers PBDTTPD, PCDTBT, and PTB7, within the following OPV device stack: ITO/(interfacial layer or layers)/BHJ/LiF/Al/Mg. It was found that modification of the ITO/PEDOT:PSS electrode with PBDTTPD-COOH resulted in statistically significant increases of power conversion efficiency for the PBDTTPD- and PCDTBT-based donor polymer:PC71BM BHJs, but not for the PTB7:PC71BM BHJ. Ultraviolet photoelectron spectroscopy (UPS) enabled determination of the respective energy level diagrams for these three different polymers relative to the ITO/PEDOT:PSS/PBDTTPD-COOH electrode, and revealed no injection barrier in all three polymer/substrate pairs. The observed differences of efficiency were not, therefore, electronic in origin. ToF-SIMS depth profiling and detailed experiments to determine surface energies strongly suggested that the greatest factor influencing device performance was a significant change of the local composition of the BHJ at this interface. When favorable accumulation of the donor polymer at the PEDOT: PSS/interfacial layer was observed, the result was higher OPV device efficiencies. These results suggest that for each BHJ, the surface energies of the electrodes need to be carefully considered, as they will influence the local composition of the BHJ and resulting device performance.