Abstract

Nearly a decade ago researchers started integrating organic electro-optic (OEO) materials into silicon photonic, plasmonic, and metamaterial devices structures. OEO materials have been integrated into vertical and horizontal slot silicon waveguide structures with dimensions as small as 25 nanometers. Sub-1 volt operation has been demonstrated for a variety of device structures and operation to tens of gigahertz has also been demonstrated. The fundamental response time of OEO materials is on the order of 10 femtosecond and all-optical modulation has been demonstrated to ten terahertz. The EO bandwidth of devices is typically limited by the resistivity of electrodes use to achieve modulation. Optical loss of OEO materials at telecommunication wavelengths is typically on the order of 1-2 dB/cm. Hybrid silicon photonic, plasmonic, and metamaterial devices incorporating OEO materials permit dramatic reduction in device dimensions and sub-wavelength confinement of light has been routinely demonstrated for these three classes of devices. Thermal and photochemical stability of OEO materials has been improved by orders of magnitude through lattice hardening protocols. The major concern with respect to the utilization of OEO materials for fabrication of hybrid devices has been processability and in particular the introduction of electro-optic activity through electric field poling. While single crystals of OEO materials can be grown in some cases and OEO materials can be prepared by sequential synthesis/self assembly techniques in other cases, virtually all hybrid devices have involved electric field poling of melt processable OEO materials. This approach is both rapid and cost effective but the low dielectric breakdown threshold of silicon and the poor electrical conductivity of doped silicon has limited the magnitude of poling voltages that can be applied to OEO materials. Instead of hundreds of volts/micron applied in the poling of thin (e.g., 1-3 micron) films of OEO materials, poling voltages in hybrid devices are typically limited to a few tens of volts/micron. Thus, instead of the 200-500 pm/V electro-optic activities realized in triple stack all-organic EO modulators, the effective electro-optic activity of OEO materials in hybrid device architectures is typically 30-100 pm/V. The exception to this rule is photonic crystal slotted silicon device structures where effective electro-optic activity as high as 1000 pm/V has recently been demonstrated but this high electro-optic activity is due to the device architecture and not due to efficient poling of the OEO materials. Realizing that poling fields applied to silicon-based devices will be very limited, a major focus of OEO materials research has been to develop materials with dramatically improved poling efficiencies and which also exhibit low optical loss, excellent stability, and desired processability. New time-dependent density functional theory (TD-DFT) and coarse-grained statistical mechanical (e.g., Monte Carlo and molecular dynamics) computational methods have played important roles in guiding the development of new materials appropriate for integration with silicon photonics, plasmonics, and metamaterial architectures. Three classes of OEO materials have been defined by theoretical analysis: (1) organic chromophore-composite materials, (2) organic chromophores covalently incorporated into dendrimers, polymers, and dendronized polymers, and (3) matrix-assisted poling macromolecular materials. The first class of materials has traditionally been the workhorse materials for prototype device fabrication and such materials are still widely used including for the fabrication of hybrid devices. The best possible poling efficiency that can be obtained with these materials approaches Langevin behavior where the acentric order parameter, <cos³θ>, approaches the poling energy divided by 5 times the thermal energy. Chromophore shape is the critical parameter that can be engineered to optimize poling efficiency. For the second class of materials, the restrictions placed by covalent bond potentials, in addition to shape considerations, define poling efficiency and the best possible performance that can be realized is that of chromophores behaving as independent (non-interacting) particles. For these first two classes of materials only 2-3% of the fundamental molecular hyperpolarizability of organic chromophores is effectively translated to macroscopic electro-optic activity. Thus, instead of electro-optic activity significantly in excess of 1000 pm/V, activities of tens of pm/V are typically realized. In the third class of materials, specific spatially-anisotropic interactions are synthetically-incorporated into chromophore structures to promote acentric assembly leading to reduced lattice dimensionality and dramatically improved acentric order and thus to improved electro-optic activity. Such interactions can be dipolar, quadrupolar, or ionic in nature. These interactions also improve material processability, stability, and homogeneity leading to reduced optical loss and improvement in other auxiliary properties. Quantum and statistical calculations suggest a template for the systematic improvement of organic electro-optic materials and potentially to the engineering of melt processability OEO crystalline materials with desired crystal structures and shapes. Laser-assisted electric field poling (LAEFP) has been experimentally and theoretically investigated as a route to further exploitation of matrix-assisted poling materials. Indeed, amorphous OEO materials have been processed into perfectly order electro-optic materials in a matter of minutes exploiting. Such experiments provide additional insights into the role of introduced spatially-anisotropic intermolecular interactions in defining molecular order and cooperativity. A variety of new and modified techniques have been employed to provide experimental definition of the role of various molecular interactions in defining poling-induced acentric order. These techniques include variable angle polarization referenced absorption spectroscopy (VAPRAS), variable angle spectroscopic ellipsometry (VASE), shear modulation force microscopy (SM-FM), intrinsic frictional analysis (IFA), and broad band dielectric relaxation spectroscopy (DRS). These experimental methods, together with a variety of electro-optic and molecular hyperpolarizability measurements, permit definition of even and odd order parameters, lattice dimensionality (defined by the ratio of odd and even order parameters), molecular cooperativity, and material phase transition temperatures. The correlation of experimental and theoretical data has provided striking new insights into the electronic and photonic behavior of organic electroactive materials. This includes understanding factors that lead to spectra line broadening and shifts that occur in transitioning organic electroactive materials from solution to solid state materials. Indeed, recently linear optical spectra for organic nonlinear optical materials have been obtained that approximate the spectra observed for the same chromophores in solvents of comparable dielectric constants. This is a critical advance for controlling optical loss.

© 2013 Optical Society of America