Controlling crystal evolution on the nanoscale: Polymers

Abstract

Organic semiconductors have applications in light-emitting displays and printable electronic circuits. However, they exhibit strong trapping of electrons but not holes. Organic fieldeffect transistors (FETs), therefore, typically show p-type but not n-type conduction (other than for special highelectron-affinity and narrow-band-gap organic semiconductors, which are susceptible to quasi-irreversible doping). A team led by Richard H. Friend of the University of Cambridge, together with researchers from the National University of Singapore and Singapore’s Institute of Materials Research and Engineering, has shown that the electron trapping occurs at hydroxyl groups at the interface with the FET gate dielectric (e.g. silanol, SiOH, for commonly used SiO2) [Chua et al., Nature (2005) 434, 194]. The researchers demonstrate that the use of alternative gate dielectrics, including polyethylene, parylene, and poly(methyl methacrylate), allows n-FET conduction. But the team also shows that a preferred choice is a crosslinkable divinyltetramethylsiloxanebis(benzocyclobutene) derivative (BCB), since it can provide a high-quality, completely hydroxyl-free interface. An advantage of BCB is that it has a high dielectric breakdown strength of >3 MV cm-1. Furthermore, it can be solution-cast to form the ultrathin films needed for practical low-gate-voltage plastic transistors. Unexpectedly, it was discovered that the use of BCB can yield ambipolar (i.e. both p-channel and n-channel) FET conduction in most conjugated polymers. Specifically, n-FETs show low threshold voltages (VT 105). High electron fieldeffect mobilities of 10-3-10-2 cm2 V-1 s-1 – at least as high as hole mobilities – have been measured in polymers based on poly(fluorene) and dialkyl-substituted poly(p-phenylenevinylene), all in the unaligned state. Proper chain alignment (e.g. in the liquid-crystalline phase) could yield electron mobility values approaching those for hole mobility in these materials with a SiO2 dielectric. The researchers say that ambipolar behavior should greatly extend the range of suitable materials for organic complementary metal-oxide semiconductor (CMOS) circuits. The findings should also be relevant to cases where molecular semiconductors are in contact with inorganic oxide surfaces, e.g. in photovoltaic applications and single-molecule devices. Mark Telford New dielectrics open gate for n-type organic field-effect transistors The study of most crystallization processes using X-ray diffraction or spectroscopic tools is limited until the crystal reaches a critical size of many microns – too late to determine the relationship between crystal growth and environmental parameters. Scanning probe microscopy allows in situ visualization, but typically of structures growing randomly on a surface from a bulk solution. Now, Chad A. Mirkin and coworkers at Northwestern University have developed a method for initiating site-specific crystal nucleation of a polymer, controlling growth in a serial manner, and monitoring progress from nanoscopic seeds to macroscopic crystals as a function of environmental conditions [Liu et al., Science (2005) 307, 1763]. They use dip-pen nanolithography (DPN), which uses a coated atomic force microscope (AFM) tip to transport adsorbate to a surface in a controlled manner. A Si AFM tip, coated with poly-D,L-lysine hydrobromide (PLH), forms triangular prisms of PLH when raster-scanned in tapping mode over an 8 μm x 8 μm region of a freshly cleaved mica substrate at room temperature. There are just two orientations, differing by 180°, independent of scan direction. This indicates oriented epitaxial growth with respect to the pseudo-hexagonal lattice of the mica surface. More scans or a longer tip-substrate contact time increases crystal thickness and edge length. Edge length grows at a faster rate than thickness, partly because of the strong electrostatic interaction between the PLH molecule and substrate, which accelerates crystal growth along its edges. Small changes in environment (i.e. temperature, humidity, or pressure) can have enormous effects on the propensity to crystallize and the ultimate crystal morphology. In subsequent scans at 35°C, cubic-shaped features emerge at the corners or edges of the prisms. The morphological change is very reproducible. Mirkin says that they can determine why a crystal forms a particular shape as a function of environmental conditions. Lower humidities favor the formation of smaller crystals, down to five orders of magnitude smaller than those observable by single-crystal X-ray diffraction, allowing observation of otherwise undetectable morphological changes. This may allow systematic determination of the optimum growth conditions for crystals that are normally difficult, or impossible, to grow. Since DPN can be a massively parallel tool, combinatorial approaches could identify conditions for initiating a particular type of crystal growth for a given set of target molecules. Obvious candidates are proteins and organic macromolecules. Mark Telford