ABSTRACT
The development and application of organic electro-optic (OEO) materials can be divided into two time periods. The first time period was addressed by the chapter on OEO materials in the 2006 edition of Handbook of Conducting Polymers [1] and largely focuses on the development of all-organic, polymer-based electro-optic modulators that were primarily designed to compete with discrete lithium niobate devices [1–6]. The OEO materials at the time were mainly chromophore-polymer composites and traditional polymers (such as polymethylmethacrylates, polycarbonates, and polyimides) containing covalently incorporated chromophores [1–6]. In both of these cases, chromophores, with significant molecular first hyperpolarizability, were incorporated at low number densities (e.g., 20% or less chromophore loading by weight). This earlier period of time reflects guidance in the design of materials from quantum [1, 5, 7] and statistical mechanical [1, 8–13] methods that were largely focused on non-interacting chromophores in the case of quantum mechanical methods [1, 5, 7] and on moderately dilute chromophore materials in case of statistical mechanical methods [1, 5, 8–13]. The second time period dates from roughly 2005 and focuses more on the integration of OEO materials with silicon photonic, plasmonic, metamaterial, and photonic crystal device architectures. Competitive technologies are now largely those based on semiconductor materials utilizing the quantum confined Stark effect (QCSE), free carrier dispersion (FCD), or the Franz-Keldysh effect (FKE); although as will be discussed later in this chapter, advances in material processing concepts have permitted lithium niobate to remain competitive. This more recent research focus also marks an increased reliance on advanced multi-scale (correlated quantum and statistical mechanics) theoretical methods to guide the development of new OEO materials [5, 14–36] and to consider OEO materials with high chromophore number densities existing in condensed phases (solids or melts at their material glass transition temperature during electric field poling). In this second phase of development, dramatic improvement in electro-optic device performance has been achieved, e.g., the voltage-length (UπL) performance of devices has been improved by nearly a factor of 1,000 (from 24,000 V-µm to 40 V-µm) from the earlier period, facilitating development of devices with footprints approaching 1 µm2, energy efficiencies on the order of a femtojoule/bit (for digital signal processing), bandwidths of greater than 100 GHz (in some cases, greater than 1 THz), in-phase-quadrature (IQ) modulation greater than 100 GBd over kilometer distances, and exceptional signal linearity (bit error ratio (BER) in the case of digital information processing and spurious free dynamic range (SFDR) for analog information processing) [37–50]. There has also been a dramatic evolution in the type of OEO material considered. State-of-the-art OEO materials are now high chromophore number density materials that explicitly incorporate intermolecular interactions that improve poling efficiency and other physical properties [5, 14, 23–26, 34–36, 51–53]. Moreover, theoretical analysis of effects associated with material interfaces in devices of nanoscopic dimensions has become important for understanding poling efficiency and device performance [34–37, 54, 55]. As the molecular first hyperpolarizability of chromophores was improved in the decade preceding 2005, the dipole moments of chromophores were also increasing. As can be seen from the theoretical simulations (based on coarse-grained Monte Carlo statistical mechanics [15, 17, 18, 34–36]), optimum electro-optic activity is predicted to occur at low chromophore loading for low to modest dipole moment chromophores. However, as chromophore dipole moments increase, optimum electro-optic activity is predicted to occur at higher chromophore number density (which is counter intuitive). As demonstrated in recent publications [17, 18, 36, 37], the height of the maximum at high chromophore number densities observed for the latest chromophores is even more dramatic in favoring high number density materials. As dipole moments are increased further (i.e., to zwitterionic ground state chromophores [5, 34–36]), simulations predict that poling-induced electro-optic activity will again approach zero. The work of Tobin Marks and coworkers [56] on twisted bridge chromophores appears consistent with this observation. In addition to the fact that optimum electro-optic activity is now observed at high number density for the highest hyperpolarizability neutral ground state chromophores, it is also noted that optimum electro-optic activity (or more precisely, the product of chromophore number density and acentric order) is observed for a narrow range of dipole moments. As has been recently demonstrated [34–36], quantum mechanical simulations have shown that this range of dipole moments also yields optimum molecular first hyperpolarizability (β). The details of chromophore shape have a dramatic impact on simulation results and observed experimental performance [15, 17, 18, 34–36]. As with earlier calculations, more recent simulations demonstrate that increasing the waist of chromophores leads to improved poling-induced acentric order but as this modification also leads to decreasing chromophore number density, a maximum is observed for electro-optic activity with adding increasing bulk to the waist of chromophores. Moreover, the effect depends on whether chromophores are linear (straight) or bent with the largest maximum electro-optic activity predicted for linear chromophores. Also, interactions that influence the matrix dimensionality (rotational freedom) felt by chromophores during electric field poling can significantly influence maximum realizable electro-optic activity with lower dimensionality leading to improved electro-optic activity.
