New types of on-fiber active devices may help create a fully all-fiber network and quicken the paradigm shift necessary to meet the demands of 5G.
MAHMOUD A. EL-SHERIF, DREXEL UNIVERSITY (RET.) AND PHOTONICS INC.
To take advantage of 5G technology,
networks must be capable of moving huge amounts of data at unimaginable speeds and extremely broad bandwidth. 5G networks should enable downloads at a rate of more than 30 gigabytes per second (GB/s), which is more than 1000× faster than the current 4G. Unfortunately, available technologies are not qualified to unleash these capabilities. Available communications networks are hindered by the integration of optical devices used to manipulate signals. Such devices include modulators, switches, couplers, and dense wavelength division multiplexers (DWDMs). Additionally, integrated optical devices are constructed of rectangular waveguides that connect to cylindrical optical fibers. This design results in high insertion loss, coupling problems, low speed, and problems due to polarization dependency. For the past three decades, significant research has been done to overcome these limitations, with only moderate success. Incremental improvement is no longer acceptable for the coming 5G networks. There is a strong need for a paradigm shift in communications technology and networks.
Recent research indicates that new types of on-fiber devices (OFDs) can be built directly onto the core of optical fibers. This approach takes advantage of nanoprocessing — the right tool for low-cost manufacturing of OFDs — and newly developed electro-optic (EO) polymers. By using OFDs, optical signals are able to transmit seamlessly within an optical fiber core — from the sending point to the receiving end — and eliminate the need to cut the optical fiber link or introduce integrated optical devices within the network. With OFDs, the optical fiber in any network will serve as both a device and a communications link. Furthermore, its fibers are expected to manage ultrahigh speed (up to 140 GB/s and more) with extremely broad bandwidth. Fibers with these characteristics can overcome existing limitations within communications networks and may help realize, for the first time, truly all-fiber networks.
Available technology
In optical communications networks, signal manipulation such as modulation, coupling, switching, or multiplexing is performed by an integrated optical device. Introducing this integrated optical device to any communications network requires cutting an optical fiber cable/link to connect it. However, since most of these integrated devices are constructed of
rectangular waveguides, additional components are necessary for efficient coupling to the cylindrical optical fiber. Even with the use of these additional components, communications networks suffer high insertion loss and limited speed, and they are polarization dependent. The effect of this dependence is particularly strong on high-speed digital communications.
Integrated lithium niobate (LiNbO3) modulators, for example, which are used as the backbone in most telecommunications networks, suffer similar limitations. Current research seeks to replace the LiNbO3 material with high-speed EO polymers that are less costly and have a much lower refractive index than LiNbO3 for better coupling to optical fibers. However, because of the rectangular structure of the optical waveguide, the device’s polarization dependency is still a major problem, and it limits high-speed application in digital communications networks (Figure 1a).
In test results (Figure 1b), when a signal of a triangular wave shape (upper) is applied as the modulating signal at 1.0 kHz, the speed limitation is obvious; the resulting modulated waveform is a distorted triangular signal (lower) with double peaks, limiting the speed of digital applications1. The modulated signal cannot follow the shape of the modulating signal. Instead of having a single peak, as in the modulating triangle signal, a double peak is shown in the modulated signal. Therefore, this type of modulator can’t be used in high-speed digital communications. Even when the modulator is positioned near the light source, the polarization is still a problem.
Figure 1. A schematic of an integrated electro-optic (EO) polymer modulator with a rectangular waveguide (top), and the test results (bottom), showing the modulating signal (upper) and the modulated signal (lower). Courtesy of Reference 1.
New approaches
By building EO devices directly onto the surface of the optical fiber’s core, in a cylindrical symmetric and uniform structure, the device’s polarization dependence, as well as coupling problems, can be eliminated in the fiber network (Figure 2). The device design is based on modifying a small section of an ordinary optical fiber. The passive homogenous cladding material can be replaced by a multilayer nanocoating, which includes a layer of EO polymer material. The newly modified multilayer cladding must surround the fiber core in a symmetric and uniform 360°. This coating may be applied at any location along the optical fiber link in any all-fiber network. The advantages of using this method include reduced cost, higher signal-to-noise ratios, and ultimately more efficient ultrahigh-speed communications.
Figure 2. A schematic showing the difference between an existing integrated optical device (top) and a newly developed OFD (bottom). Courtesy of M.A. El-Sherif.
Method of modification
The method of modification takes advantage of nanoprocessing and newly developed EO polymers2. The modified section can be in the range of 200 to
300 µm. The modification is performed by mechanically stripping the fiber jacket and then etching the passive cladding material by a wet etching technique. The modified section is then coated, in 360°, with a multilayer cladding. The cladding includes a thin layer of an EO polymer3 that is sensitive to electromagnetic fields and coated by spin coating. Then, in situ poling is applied to the polymer to improve and maximize the EO properties3. This layer of EO polymer is sandwiched between two cylindrical metallic inner and outer electrodes. A plasma-enhanced deposition technique in a cylindrical structure is used to coat the electrodes. Then, the jacket material is applied on the top of the outer electrode. The design of the jacket includes two exposed metallic rings, which are connected from inside with the inner and outer electrodes.
Figure 3 shows the major processing steps.
Figure 3. A schematic showing the major design and manufacturing steps: an ordinary optical fiber
(core, clad, and jacket) (a); the fiber after stripping the jacket, etching the cladding, and applying the mask
for coating the inner electrode (b); after coating the inner electrode and before removing the mask (c);
after removing the old mask and applying the new, then coating the EO polymer layer (d); after applying
the new mask and coating the outer electrode (e); and after applying the new jacket, with two metallic rings connected from the inside with the two electrodes (f). Courtesy of M.A. El-Sherif.
Before applying each coating layer, the surface of the modified section is treated for better adhesion of the next coating, and a specific mask is applied based on the geometry/design of the next coating layer. This mask will be removed after coating this specific layer to prepare for the next mask and next coating4-7. The design of the multilayer modified cladding will be different from one device to another based on the application. Such devices/applications are EO modulators, switches, tunable couplers, and tunable DWDMs4-7.
In the presence of a modulating electronic signal applied to the device electrodes, an electromagnetic field is generated between the two electrodes, resulting in a uniform radial changing of the polymer’s refractive index. The results demonstrate a uniform modulation of optical signals propagating within the fiber core, regardless of the polarization direction of the optical signals. Thus, the problem of polarization dependence is eliminated. The optical signals will continue within the same fiber core from one device to another in the fiber network, with minimal insertion loss.
Testing and discussion
Modulators have been manufactured and packaged with tools available in the laboratory (Figure 4a). Their dimensions are 1 × 2 × 0.4 in. By using the proper packaging tools, the devices’ size can be reduced by 70% to 80% and can also be in a cylindrical shape with about a 0.5-in. diameter and a 1- to 1.5-in. length. The packaging process includes the integration of a microwave socket connected from inside with the two metallic/
electrode rings. For fiber mechanical protection, two black rubber sleeves were used to cover the exposed fiber ends.
Several devices have been tested and various types of modulating signals
(sinusoidal, triangle, and digital) were used. Test results were encouraging because the modulated signals were identical to the applied/modulating signals (Figure 4b) for triangle modulating signals. This type of modulating signal was selected to prove the high quality of OFDs as polarization-independent devices, unlike integrated EO modulators, as explained earlier in Figure 1.
Figure 4. A prototype EO modulator (top), which was tested under the application of triangle 8.1-kHz waves, with the modulating electronic signals shown in black. The modulator output, or the modulated signals shown in red, was detected by a Michelson interferometer (bottom). Courtesy of M.A. El-Sherif.
Based on recent advancements in EO polymers, OFDs have a demonstrated capability of managing up to 140 GB/s speed. This type of technology is ready for application in EO modulators, tunable
couplers, tunable 2D switches, and tunable DWDMs, and may help open the door for the next-generation 5G communications networks.
Acknowledgment
The author gratefully acknowledges that the EO polymers used in most of the developed and tested modulators were provided by Alex Jen and his research group, including Antao Chen and Jingdong Luo, of the University of Washington.
Meet the author
Mahmoud A. El-Sherif, Ph.D., has 50 years of experience in academia and industry and is a former Drexel University faculty professor (1989-2007). During his tenure, he founded and directed Drexel’s Fiber Optics and Photonics Manufacturing Engineering Center (FOPMEC), which was named a Center of Excellence. In industry, he served as CTO, president, and CEO of Photonics Inc.; email: [email protected].
References
1. F. Qiu and S. Yokoyama (Aug. 22, 2016). Efficiently poled electro-optic polymer ofds. Opt Express, Vol. 24, No. 17, pp. 19020-19025.
2. Mahmoud A. El-Sherif (2019). Nano
materials application in development of
novel class of fiberoptic modulators.
Nanotechnol Adv Mater Sci, Vol. 2, No. 2, pp. 1-4.
3. H. Ma et al. (2002). Polymer-based optical waveguides: materials, processing, and devices. Adv Mater, Vol. 14, No. 19,
pp. 1339-1365.
4. M.A. El-Sherif (Dec. 12, 2006). On-fiber tunable coupler and high speed switch for telecommunication applications. U.S. Patent No. 7149391. Washington, D.C.: U.S. Patent and Trademark Office.
5. M.A. El-Sherif (June 5, 2007). On-fiber
microwave modulator and high speed switch for telecommunication applications. U.S. Patent No. 7228012. Washington, D.C.:
U.S. Patent and Trademark Office.
6. M.A. El-Sherif (May 13, 2008). On-fiber
microwave modulator and high speed switch for telecommunication applications. U.S. Patent No. 7373024. Washington, D.C.: U.S. Patent and Trademark Office.
7. M.A. El-Sherif (Aug. 12, 2014). On-fiber tunable Bragg gratings for DWDM
applications. U.S. Patent No. 8805136. Washington, D.C.: U.S. Patent and
Trademark Office.