Photonic integrated circuits (PICs) are an area of growing research and are now present in many applications. Integrated photonics change the scaling laws of information in communication systems and offer architectural choices which combine photonics with electronics.
A team of Researchers from Europe have now developed a silicon photonics multifunctional processor core composed of a silicon integrated hexagonal waveguide mesh for use in various applications.
Photonic integrated circuits (PICs) adopt both the properties of optical waveguides with other desirable features. Commonly, PIC waveguides consist of low propagation losses, an absence of diffraction, high power confinement, low crosstalk and an immunity to electromagnetic interference which are often combined with a small footprint, compact, stable, reduced power consumption and low-cost properties.
PICs are usually specifically designed and tailored to a given application, where the internal chips are optimally designed to perform specific functionalities. However, this has led to long development times as a result of a large number of design and fabrication iterations.
A recent approach to PICs has been inspired by electronic Field Programmable GateArrays and PICs have begun to incorporate a programmable photonic processor where common hardware is implemented using a 2D photonic waveguide mesh. Such circuits have been able to realize different functionalities through a series of programs.
The Researchers from Europe have now produced a silicon multipurpose processor core which utilizes an integrated hexagonal waveguide mesh, upon which the mesh was composed of 7 hexagonal Mach Zehnder Interferometers (MZI) waveguide cells.
The device itself was fabricated at the Southampton Nanofabrication Centre using e-beam lithography to create silicon overlayers and buried oxide layers to create grating couplers. Resist stripping and dry etching followed this to create the optical waveguides from the grating couplers.
Photolithography was employed to define isolation trench openings and deep dry etching was utilized to etch through all the layers. The Researchers deposited a metal layer on top of the device, followed by further photolithography and dry etching steps to create electrodes. The layered material was then deposited onto a printed circuit board (PCB) and electrical connections were then created using a wire bonding process.
The Researchers tested the optical properties using a static characterization technique which employed a tuneable laser (ANDO AQ4321D) and an optical spectrum analyzer (ANDO AQ6217C) to characterize the differential path length, propagation losses, cascaded bends structures, bend losses, the cascaded and coupled MMI structures and the MMI insertion losses and bandwidth.
The tuneable basic units used in the device were: 3 Keihtley2401, 13 Thorlabs LDC8010 and 2 TECMA 72-2535 current sources.
The device produced two waveguides which could be coupled or switched through an MZI. The waveguide mesh was found to be configurable and could demonstrate 21 different functionalities, ranging from simple single-input/single-output FIR filters, to optical ring resonators (ORRs), coupled resonator waveguides (CROWs), side-coupled integrated spaced sequences of optical resonators (SCISSORs), ring loaded MZIs and multiple-input/multiple-output linear optic 2 × 2, 3 × 3, and 4 × 4 transformations including Pauli Matrices and a C-NOT gate.
In comparison to devices which are designed with one specific application in mind, this photonic processor can be used in a wide variety of applications by just using a single chip. Compared to application-specific devices, such photonic circuits provide flexible and adaptive topologies and circuit parameters.
These multipurpose photonic chips show a great potential for commercial production and real-world applications due to their compatibility with CMOS technologies, high production volume potential and low cost on a large scale. These multipurpose circuits have the potential to be used across a wide range of fields, including in communications, chemical and biomedical sensing, signal processing, biophotonics, photonic switching, multiprocessor networks and quantum information systems.
“Multipurpose silicon photonics signal processor core”- Pérez D. et al, Nature Communications, 2017, DOI: 10.1038/s41467-017-00714-1