Aggregation in Semicrystalline Organic Thin Films
- D.T. Duong

While the performance of semiconducting polymers has approached that of amorphous silicon, it is difficult to comprehensively describe the solid state microstructure due to the inherent materials' complexity. For inorganic semiconductors such as silicon or GaAs, the relationship between electronic properties (i.e., charge carrier density, carrier mobility) and structural properties (i.e., atomic structure, doping density) is well understood. This is not the case for organic semiconductors and the physical structure through which charge carriers propagate is only partly understood. This lack of knowledge on film microstructure makes it difficult to rationally design new polymers: Tailoring the molecular structure is not sufficient. The nature of the crystalline structure and the fraction of ordered regions in the film also affect electronic performance. We must be able to understand the film forming process, characterize the overall morphology--with particular attention paid to the structure of ordered and disordered regions--and understand how charges propagate across the proposed microstructure.

Figure 1: Proposed mechanism for P3HT solidification. Following the establishment of a solvent sheet by spin casting, an ultrathin interfacial layer of edge-on aggregates forms at the substrate-polymer interface (a). As the solvent evaporates, more edge-on aggregates nucleate off the wetting layer (b), face-on aggregates nucleate from either the polymer-air interface or in the bulk (c), and aggregates can self-assemble into lamellar-stacked crystallites (d). Once the film has dried, the final morphology consists of a mostly edge-on interface with a mixture of both face-on and edge-on aggregates and crystallites in the bulk (e).

We have shown that the microstructure of P3HT thin films consists of an ultrathin, interfacial layer of edge-on aggregates covered by a more disordered, face-on bulk film. We propose a mechanism for aggregate nucleation, crystallite assembly, and film formation. We conclude that efficient transport occurs mostly within the observed interfacial layer and requires good interconnectivity between well-ordered aggregates. Our findings therefore provide a coherent picture of the nanoscale morphology for P3HT thin films and can be used for the general design of materials for electronic devices. The ability to nucleate well-ordered, extended π- stacked aggregates off the substrate-film interface, for instance, is more important for FET applications than the inherent crystallinity of a particular polymer. These findings should also serve to change how researchers approach issues of microstructure-dependent charge transport in semiconducting polymers.

Fabrication of Organic Semiconductor Micropatterns by Capillary Force Lithography
- P.S. Jo

Soluble organic semiconducting materials have attracted much attention because they are easily processable for low-cost and large-area devices. Among organic semiconductors, conjugated polymers are suitable for solution processes such as spin-coating and dip-coating because of their good solubility in organic solvents and uniform film-forming properties. Polymers however have intrinsic entropic barriers to achieving highly-ordered packing and high crystallinity.

Figure 1: A schematic illustration of the micropattern and OFET fabrication processes: A pre-patterned PDMS stamp (p-PDMS stamp) and drop-cast droplets of the C8-BTBT solution on a substrate are prepared. A p-PDMS stamp is placed in conformal contact with the droplets of C8-BTBT solution, and thermally annealed at 50 °C for 4 h to solidify the solution. After the stamp is carefully detached, C8-BTBT line patterns are generated. Au is thermally evaporated to form the source and drain contacts of OFETs.

Films of solution-processable small molecules on the other hand may reach a high degree of crystallinity and provide sufficient overlap of π-orbitals to attain outstanding electrical performance as measured by carrier mobility. In particular, single crystals made of small molecules are desirable not only because they can be used in very high performance organic field effect transistors (OFETs), but also because they allow to investigate the intrinsic electrical properties of these materials. Nevertheless, there are still problems to be resolved for the practical utilization of single crystal organic semiconductors as active layers in OFETs. Since during solidification the nucleation of small molecules is difficult to control, it is almost impossible to fabricate single crystals that are uniform in size and shape, which causes property variations among devices even when fabricated in the same batch.

To fabricate more uniform active layers, recent work has focused on controlling the morphology of active layers by defining nucleation locations and inducing directional growth of small molecules. These methods however still need relatively complex tools, many process steps, or have limitations in controlling crystal shapes. Here, we fabricated strongly textured organic semiconductor micropatterns made of the small molecule, dioctylbenzothienobenzothiophene (C8-BTBT) using a simple method based on modified capillary force lithography (CFL) (Fig. 1). Nucleation and directional growth of small molecules via CFL was investigated with X-ray diffraction, optical and electron microscopy. As a result, we demonstrate high-performance OFETs based on strongly oriented small molecule micropatterns. By using this technique, not only is the C8-BTBT line pattern generated over a large area, but also the dimensions of the pattern replicate those of the p-PDMS stamp (Fig. 2)

Figure 2: Dimensional control of C8-BTBT line patterns: (a, b) OM and AFM images of 10 μm wide, 1.8 μm thick C8-BTBT lines spaced 20 μm apart.

Lanthanide-based Upconverting Nanoparticles
- M. Wisser, D. Wu

Although recent innovations in cell architecture and fabrication techniques have yielded substantial improvements, the efficiencies of single-junction solar cells remain fundamentally limited by transmission of sub-band gap photons. Photon upconversion (UC) represents a promising approach toward overcoming this efficiency limit; by absorbing sub-band gap light and upconverting it to energies above the solar cell’s band gap, an upconverting layer can significantly boost the efficiency of any single-junction cell by enabling it to utilize photons it would otherwise waste. This process is depicted schematically in Figure 1a.

The specific system which we study consists of Yb3+ (which absorbs 980-nm light) and Er3+ (which emits upconverted, visible light) housed in a NaYF4 matrix. This UC system is shown in action in Figure 1b. While this system has emission and absorption spectra well-suited to current photovoltaic technologies, its low quantum efficiency (~0.3%) currently prevents the material from attaining industrial relevance. Our group is actively exploring routes to boost the efficiency of this UC system and elevate the technology from a largely academic pursuit to a commercial breakthrough.

Figure 1: a) Schematic depicting proposed device architecture; note that the cell and upconverter need only be optically coupled rather than electrically as in multijunction cells (b) Dispersion of NaYF4:Er,Yb nanoparticles illuminated with near-infrared light and emitting green light

In any Er,Yb-based upconverter, the UC efficiency of the system is related to the structure of the host matrix. More specifically, the distances between the Er3+ and Yb3+ ions as well as the crystal fields which the ions experience (both of which are functions of the host lattice geometry) strongly influence the quantum efficiency of the system. Our goal is to utilize pressure to sweep this parameter space in a near-continuous fashion so as to determine the ideal lattice geometry for Er,Yb-based upconversion.

Figure 2 demonstrates how the UC emission spectrum varies as a function of applied pressure. As seen in Figure 2a, pressure-induced lattice distortion causes the upconversion luminescence to fall off rapidly, suggesting that a reduced average Er-Er separation (which increases the likelihood of undesirable cross-relaxation between Er3+ ions) is the dominant effect. However, Figure 2b shows that the spectral positions of the UC emission peaks shift with pressure, evincing orbital energy shifts caused by changes to the crystal fields surrounding the ions. Thus, it is clear that the host matrix unit cell can be tuned via pressure application. Given the efficiency decreased observed when subjecting the material to compression, complementary experiments are currently underway to characterize how the system behaves in tension.

Figure 2: a) Emission from NaYF4:Er,Yb particles when excited with 980-nm light and subjected to variable pressure (b) Emission peak shifts as a function of pressure; each line maps the position of one of the five green or two red peaks relative to its value at ambient pressure

Nanostructured ZnO for solution-based transparent electrodes
- S. Mehra

Transparent conducting oxides have gained great industrial importance. The popularity of flat-screen technology (television, computer, phones, i-pods,…) and the recent greentech boom -which put photovoltaic modules at the forefront of the quest for efficient renewable energy sources- have greatly increased the use of Indium Tin Oxide (ITO). As a result, the price of In has recently surpassed that of silver. It is widely believed that the price volatility of In will become an extremely detrimental factor in the fabrication cost of photovoltaic cells. We are currently seeking to replace ITO with a low-cost materials set that is compatible with large-area, low-cost processing techniques, such as printing.

Figure 1: Left: ZnO nanoparticle synthesis, without surfactant Right: ZnO nanowire synthesis, with surfactant

Zinc based oxides, can be easily processed from solution in order to produce thin, homogeneous films that can be used as electrodes in the aforementioned applications. Because of the abundance and low toxicity of Zn, these materials are a very attractive replacement for ITO. Pure ZnO is an intrinsic semiconductor but it can be made conductive by doping it with Al, Ga or F. Several methods have been described to form polycrystalline, highly doped ZnO, the most successful being Chemically Vapor Deposition (CVD), sputtering and spray pyrolysis of Zn based salts. These techniques, however, have many disadvantages, such as the use of expensive vacuum equipment and high processing temperatures, which makes them incompatible with flexible substrates. Moreover, effective doping can be problematic due to the high reactivity of Al with residual oxygen, preventing its uniform incorporation into the ZnO lattice.

In our group, we use a colloidal synthesis in an organic solvent for the synthesis of ZnO nanostructures (particles, rods and wires). Based on previous work by Yang et al.,1 we produced nanoparticles and nanowires (W~100 nm, L~1-2 μm). The size and aspect ratio of the nanostructures depends on the concentration of the Zn salt and the presence of surfactants (Fig 1). These nanostructured ZnO powders can be easily resuspended in ethanol and spincoated into very uniform films (Fig. 2).

Figure 2: a) SEM micrograph of ZnO NW thin film showing an isolated device with thermally evaporated Al electrodes. b) SEM micrograph showing NW orientation and direction of current flow on a parallel device. d) and e) SEM micrographs showing cross-section of NW thin film taken at 45° w.r.t. plane of substrate.

Quantum Dot Formation in SiO₂ through Femtosecond Laser Annealing
- W. Mustafeez

Figure 1. Waveguide writing in Silicon Rich Silicon Oxide through high repetition rate femtosecond lasers: focal volume with multiphoton absorption occuring below the surface of transparent material.

Femtosecond Lasers through their high peak electric fields offer excellent versatility especially in nonlinear process driven applications. They find usage in a wide array of areas from materials processing to driving free electron lasers. In case of materials processing, femtosecond lasers offer nanometer scale precision machining with minimal spread of damage.

This is possible due to the ultrashort pulse lengths which are below the timescale of many fundamental processes such as carrier phonon scattering. Hence for surface processing fs lasers are an excellent tool. However due to the high peak electric field intensities they also open up the possibility of subsurface material processing through multi-photon absorption. Using high numerical aperture objectives it’s possible to perform phase transformation or damage to very small focal volumes even with unamplified nanojoule pulse energies. The field of femtosecond laser written waveguides and other structures in various transparent materials has received wide interest with possible applications in on-chip interconnects, photonic crystals etc.

Within the motivation of on-chip photonics, work on the optical properties of silicon quantum dots or nanocrystals have been going on since the early 90s. It has been shown that the emission properties of nanometer scaled silicon crystals are much more favorable in contrast to bulk for development of light sources. In fact in 2000 (Pavesi et al.) the possibility of optical gain in silicon nanocrystals was shown for the first time. Most of these studies either employ ion beam implanted Si in a matrix of SiO2/SiN or low temperature Plasma Enhanced Chemical Vapour Deposition (PECVD) deposited non-stoichiometric SiO2/SiN followed by high temperature anneal for over an hour.

Figure 2. Subsurface phase transformation in 2µm thick SRSO using high rep-rate oscillator. Inset shows visible PL from oven annealed SRSO.

Our group is working on using a high repetition rate laser oscillator for creating waveguides in PECVD deposited SiOx. The high repetition rate ensures that the time between pulses is significantly below the thermal diffusion time and hence the successive pulses can cause an energy buildup, rapid annealing and finally precipitation of silicon nanocrystals, embedded in a waveguide. Since we are using a laser wavelength well below the band gap of SiO2 and a high NA objective we can precisely control the depth and in plane position of the energy disposition.

We use unannealed SiOx films for the processing. Figure 2 shows a diascopic microscopy image of a laser processed waveguide, approximately 3µm wide and 100µm long. This experiment was performed with 800nm Ti:Sapphire laser oscillator at 75MHz repetition rate and 5.3nJ of energy per pulse. The translation speed was about 200µm/sec. We used a 1.2NA water immersion objective to focus the beam onto the SRSO film. A standard coverslip was placed above the sample. The feature shown is not visible in reflected light microscopy hence is subsurface. Development of this process will open up possibilities in spatially precise placement of Si-NC which may find use in studying Si quantum dots more closely as well enable practical applications such as using the Si-NC sandwiched between metals for coupling light onto MIM plasmonic waveguides etc.