The search is on for new building blocks to increase the speed of optoelectronic components and decrease electrical power consumption. High-performance electro-optic (EO) materials have matured commercially1 and have the potential to offer solutions. These polymer-engineered materials are naturally fast, are highly responsive to electric fields, can be added to integration semiconductor platforms such as silicon photonics or indium phosphide (InP), or can be used on their own in a packaged polymer chip. While their potential advantages were recognized several decades ago2, EO polymers are qualitatively different from the semiconductor materials that the industry is accustomed to. Not surprisingly, some of the technical commercialization challenges EO polymers have faced are similar to those that OLEDs overcame nearly 20 years ago. After a wave of investment in research and commercialization efforts, there was a fiscal shift. Large and small corporations pulled back, and government funding and venture capital dried up. EO polymers are ideally positioned at the confluence of speed, chemistry, and a matured optical integration with semiconductors. Today, however, EO polymers are ideally positioned at the confluence of speed, chemistry, and a matured optical integration with semiconductors. Silicon photonics has recently reached commercial deployment, but its roadmap increasingly includes ancillary materials, such as SiGe (silicon germanium), which are necessary to enhance performance for extended wavelength photodetectors. In the latest academic research, though, EO polymers and new device designs have combined to create some of the fastest silicon photonics3. The modulator is a key component of a high-speed photonic integration platform, and the potential combination of high-performance polymer modulators with silicon photonics or InP has incredible potential in the marketplace. Electro-optic devices The modulator is a familiar electro-optic device, classically implemented in lithium niobate. It has a known electric signal that is transferred with high fidelity onto an optical signal for transmission in fiber communications. The modulator works via the linear electro-optic, or Pockels, effect; the electric field alters the refractive index of the electro-optic material, which in turn causes a phase change to any transiting optical signal. A second application of the Pockels effect is electric field sensing, which is when an unknown electric field is similarly mapped into an optical signal that can be more easily measured. This technique is particularly valuable for sensing high-frequency electromagnetic waves such as millimeter or terahertz (THz) waves. EO polymers can address both of these applications and more4,5 (Figure 1). Figure 1. Applications of the Pockels effect include electro-optic modulators and field sensors. Courtesy of Lightwave Logic. Need for speed, power, efficiency Over time, most electro-optic applications demand higher speed and sensitivity (with respect to the applied electric field), needs that are well matched to EO polymers’ strengths. In the fiber optic communications market today, 400-Gb/s optical transceivers are evolving quickly toward 800-Gb/s and 1.6-Tb/s data rates. Mobile wireless frequency bands are also moving to higher frequencies, with 5G bands needing both analog electronics in the 60-GHz band, and supporting digital electronics for 100 and 400 Gb/s data rates. The real challenge is to increase speed while decreasing power consumption. High-speed analog electronics, even more than high-speed optics, are power hungry. Fiber optic transceivers for 400 Gb/s, for example, have a footprint similar to that of a 1-Gb/s transceiver. If size is maintained and speed is increased by orders of magnitude, the ability to control power consumption is critical. With modulators and sensors that possess higher sensitivity, electrical amplifier stages could be eliminated, reducing both cost and power. Fundamental material differences When compared to competing materials, EO polymers have several key differences, making them both a promise and a challenge. Notably, the polymers exhibit intrinsically faster response. Compared to semiconductors, such as silicon and InP (commonly used for today’s optical communications), EO polymers use a different physical effect. They use the Pockels effect, which is a “pure” linear electro-optical effect. The alteration to the electro-optic material is a distortion (hyperpolarizability) to the electron clouds of the molecules without removing them from their orbitals. The response timescale is of the order of 10−13 s. In contrast, silicon works on the basis of injection/extraction of electrons, a relatively slow process (10−10 s), resulting in plasma dispersion-induced index changes. Many III-V semiconductor (InP) devices work via the electro-absorption/quantum-confined Stark effect, which is also affected by electron flow. There is another speed limitation for traveling-wave devices where the radio frequency (RF) wave and the optical signal interact as they both travel through the nonlinear material. As the signals travel at various speeds — depending on the RF dielectric constant and optical refractive index, respectively — the RF and optical signals eventually get out of phase, limiting the useful interaction length. In fact, continuing past the point when they are out of phase will actually start to undo the effect. Higher modulation frequencies get out of phase more quickly. Therefore, for a given electrode length, the phase velocity mismatch can limit the bandwidth. Polymers have better phase velocity matching, resulting in a figure of merit at least double lithium niobate’s 43 GHz-cm. The phase-matching bandwidth limit can be raised by shortening the electrode, but only at the cost of raising the drive voltage. This brings up the topic of power consumption. Another significant advantage of EO polymers is the efficiency of converting an electric field to an optical phase change. The strength of response for a given nonlinear material is characterized by its second-order nonlinear susceptibility, with r33 being the metric for the optimal case where the direction of the applied electric field and the optical electric field is the same. Typical r33 values for lithium niobate devices are in the 30- to 40-pm/V range. (The metric for the case when the electric fields are in orthogonal axes, r13, is typically about one-third of r33). EO polymer devices a decade ago started in the 50- to 80-pm/V range but have since improved to over 200 pm/V6. Chemistry as design tool EO polymers are a vast class of materials that allow many configurations to optimize for a broad range of applications. The other materials mentioned above are not only crystalline but have fixed chemical compositions. Polymers are engineered, and as such, can be modified to achieve desired optical, electrical, mechanical, and material (dielectric) properties. The active ingredients are molecules classified as chromophores. Guest-host materials are one class of EO polymer in which a chromophore is mixed with a passive host polymer. To have the Pockels effect, the material must be nonsymmetric. This is achieved in EO polymers by poling or orienting the originally randomly oriented chromophores with an applied electric field (Figure 2). Figure 2. Activation of nonlinearity by voltage poling of chromophore molecules. Courtesy of Lightwave Logic. Poling is done at elevated temperatures when the polymer is relatively soft, above its glass transition temperature, or Tg. The orientation is retained by cooling the poled material down below Tg after the molecules are in place, trapping them in a metastable state. The chromophore is designed to maximize polarizability while retaining stability. Depending on the application, raising or lowering the glass transition temperature may be required. For each chromophore, having a variety of host polymers further extends the parameter space. In the past decade, these very large molecules (and even more complex aggregate environments, including multiple molecules and the host material) have become computationally tractable, promising rapid progress yet to come over the next decade. Noncrystalline nature Unlike semiconductors and dielectrics, which have well-defined crystalline structures that restrict their properties and device orientations, polymers have no intrinsic orientation. While less obvious than the performance advantages, this property can be exploited in device designs. Note that devices made on the same wafer, or even parts of the same device, can be poled in various orientations as determined by the location of the poling electrodes. For example, the two arms of a Mach-Zehnder interferometer can be poled in opposite directions vertically to achieve push-pull operation from the same drive signal. The noncrystalline nature of EO polymers also means that integration with other materials platforms is not constrained by crystal lattice matching or the need to do wafer bonding of dissimilar materials. EO polymers can be added to a variety of other wafer-scale integrated devices, such as silicon photonics, much the same way as photoresist. Figure 3 conceptually depicts three classes of hybrid devices: silicon plasmonics enhanced by polymers3, polymer/sol-gel waveguides4, and silicon-organic hybrid slot waveguides7. Figure 3. The classes of hybrid polymer waveguides: a dielectric passive ridge waveguide coupled to a ridge polymer waveguide (a); a silicon-organic hybrid silicon slot waveguide (b); and an optically fed plasmonic waveguide (c). Courtesy of Lightwave Logic. Challenges, recent progress One of the main figures of merit that engineers have sought to increase is the nonlinear coefficient r33. The larger this parameter, the smaller the voltage required to achieve device performance, all else being equal. Figure 4 shows a typical trade-off between device speed and drive voltage for a simple traveling-wave Mach-Zehnder where the speed is limited by the length of the device. The only way to reduce the voltage while maintaining speed for this device is to increase the material’s r33 coefficient. Silicon slot and plasmonic devices have submicron guides and effective electrode spacing. As a result, they are much shorter but incur similar benefit from an increased r33. Figure 4. Reduction of required drive voltage by increasing r33. Note that bandwidth is inversely related to electrode length. Courtesy of Lightwave Logic. Developing accurate metrology for the r33 has been a challenge. There is no industry standardized test methodology or commercial instrumentation for these materials, which makes comparisons across results difficult7. Considerable effort has been put into improving metrology, and there is now confidence in the higher susceptibility values and the process to qualify new materials. The EO polymer materials and devices historically have been designed for 1.5-µm wavelengths, suitable for telecom applications. Recent progress has extended the applicability to datacom at 1.3 µm. Typical chromophores of this class have a broad absorption peak extending several hundred nanometers around 0.8 µm, causing resonant enhancement of the value of r33 even at 1.3 µm. However, prior materials with high r33 values at 1.3-µm wavelength may not be usable in devices due to high optical loss. Recently, materials that combine high r33 values with low loss at 1.3 µm have been achieved. As with any new material, reliability must be demonstrated. Although polymers and other organics are well established in passive optical applications — such as for optical lenses, optical waveguides, optical mux/demux, and so on — active polymers have additional functionality. The two additional keys to long lifetime are retention of alignment (poling) and durability of the chromophore molecule. Lifetime and thermally accelerated testing of device attributes such as Vπ (the voltage required to induce a 180° phase shift) captures both effects. Recent progress in molecular engineering has also improved both alignment and molecular stability. The significance of Tg for poling has already been described above. The further the poled material can be kept away from Tg the better the poling retention. In other words, for a given operating temperature range, higher Tg is better. At the same time, a moderate Tg may be desirable to simplify poling conditions. There also tends to be a trade-off between Tg and high r33. It is the job of the chemistry designer to manage this balance of requirements through advanced simulation, modeling, and demonstration. Note that the Tg of the guest-host mixture depends on the Tg of the host, the Tg of the chromophore guest, and interactions between the two (plasticization). Given the high device bandwidths, short electrical leads are likely needed for many applications. Another key property that chromophore engineers are now seeking to develop is the ability to survive thermal cycles for chip packaging methods such as ball grid arrays, flip-chip bumping processes, wave-soldering processes, and more. The chromophore itself has a characteristic decomposition temperature Td, which can be highly dependent on the design. Like Tg, a higher value means more stability at device operating, storage, and processing temperatures. Values approaching 300 °C have been demonstrated. Photostability is the ability of the chromophore to tolerate high-intensity light. In addition to the expected light level for optical device operation, the material may need to tolerate or be shielded from UV exposure during device fabrication. The long-term stability of the material is improved with hermetic or quasi-hermetic packaging. A key challenge currently is to evolve from traditional hermetic packages to chip-level encapsulation techniques. Note that OLEDs have similar requirements for their organic compounds, and methods have successfully been developed for those devices. Optical communications applications have reached data rates and power consumption levels at which it is generally acknowledged that further increases are becoming more difficult. The roadmap for silicon photonics is to add more materials to its toolkit, with faster modulators being a particular gap. The high speed and low power attributes of EO polymers, along with their natural compatibility of fabrication processing with silicon structures, look to be an excellent fit to meet this need. Recent progress and the prospect of further computationally aided molecular design promise continued improvement in EO polymers. Meet the authors Karen Liu, Ph.D., is vice president of sales and marketing at Lightwave Logic Inc. She has a doctorate in applied physics from Stanford University and extensive fiber optics industry experience; email: karen@lightwavelogic.com. Cory Pecinovsky, Ph.D., is director of materials development at Lightwave Logic Inc. He has led a variety of commercialization programs in polymers and liquid crystals since receiving his doctorate in organic chemistry from the University of Colorado; email: cory@lightwavelogic.com. Michael Lebby is CEO of Lightwave Logic Inc. and is a well-known technology and business leader in optoelectronics/photonics, electronics, and semiconductors; email: michael@lightwavelogic.com. References 1. M.S. Lebby (April 2020). Naturally fast and low power electro-optic polymer optical devices are ideally positioned for the next-generation internet photonics roadmap. Proc SPIE, Vol. 113640, Integrated Photonics Platforms: Fundamental Research, Manufacturing and Applications, www.doi.org/10.1117/12.2571129. 2. L.A. Hornak, ed. (1992). Polymers for Lightwave and Integrated Optics: Technology and Applications. New York: Marcel Dekker Inc. 3. W. Heni et al. (2020). Ultra-high-speed 2:1 digital selector and plasmonic modulator IM/DD transmitter operating at 222 gbaud for intra-datacenter applications. J Lightwave Technol, Vol. 38, No. 9, pp. 2734-2739. 4. Y. Enami et al. (2018). Bandwidth optimiza- tion for Mach-Zehnder polymer/sol-gel modulators. J Lightwave Technol, Vol. 36, No. 18, pp. 4181-4189, www.doi.org/10.1109/JLT.2018.2860924. 5. C.-J. Chung et al. (2018). Silicon-based hybrid integrated photonic chip for k band electromagnetic wave sensing. IEEE J Lightwave Technol, Vol. 36, No. 9, pp. 1568-1575. 6. C. Kieninger et al. (2018). Ultra-high electro-optic activity demonstrated in a silicon-organic hybrid modulator: supplementary material. Optica, Vol. 5, Issue 6, pp. 739-748. 7. R. Palmer et al. (2014). High-speed low drive-voltage silicon-organic hybrid modulator based on a binary-chromophore electro-optic material. IEEE J. Lightwave Technol, Vol. 32, Issue 16, pp. 2726-2734.