BELFAST, UK – A team of European researchers has demonstrated what could be the basis of novel photonic components. Through the exploitation of plasma oscillations, these and other plasmonic devices potentially could manipulate both electronic and photonic signals simultaneously. Thus, they could allow all-optical signal switching, modulation and selection.
Using plasmonics makes optical components smaller, in this case a plasmonic waveguide-ring resonator with a radius of 5 μm. Shown are topographical (a) and near-field optical images measured with 1530-nm (b), 1550-nm (c), 1570-nm (d) and 1590-nm (e) light sources, indicating wavelength-dependent transmission and selection. Images courtesy of T. Holmgaard, EC FP6 STREP Plasmocom (European Commission’s Sixth Framework Programme’s Specific Targeted Research Projects).
Anatoly V. Zayats, an optics professor at Queen’s University Belfast and a member of the research group, noted a specific advantage to the plasmonic approach as compared with other techniques. “With plasmonic waveguides and, in particular, dielectric-loaded plasmonic waveguides that we are researching, the bend loss is much smaller. The small bend loss allows us to achieve a smaller size of wavelength-selective components.”
Current commercial ring resonators, for example, have a radius of 25 to 300 µm. The corresponding plasmonic device would have a radius of only 5 µm – more than an 80 percent reduction.
The group reported on its work in the online editions of Optics Letters and Applied Physics Letters
in February, with researchers from Queen’s University, Aalborg University in Denmark, the University of Southern Denmark in Odense and the University of Burgundy in France involved. The effort is part of Plasmocom (polymer-based nanoplasmonic components and devices), a European Commission research project.
The investigators fabricated wavelength-selecting devices using deep-ultraviolet lithography, employing sub-250-nm light to pattern a layer of poly(methyl methacrylate) resist that they had spin-coated on a 60-nm gold film supported by a thin glass substrate. With the patterning, they constructed features that were 500 nm in size, far below the 1500 nm or so wavelengths they were interested in controlling.
Shown are scanning electron microscope (a) and topographical (b) images of a directional plasmonic coupler. Near-field optical images of the same device under various wavelengths (c-e) illustrate the wavelength-dependent output. Inset is a scanning electron microscope image of the coupling region. S = coupler separation.
They made directional couplers in one set of devices, and waveguide-ring resonators and in-line Bragg gratings in the other. In the first group, they demonstrated that a 50.5-µm-long coupler should enable complete separation of channels at 1400 and 1620 nm, an important telecom wavelength range. In the second group, they showed that a 150-µm2
resonator could act as an efficient and compact wavelength-selective filter for the same span.
Although the devices were smaller than their conventional counterparts, they did exhibit somewhat inferior performance. A traditional ring resonator, for example, has a 20-dB contrast, better than the 13 dB exhibited by the plasmonic devices. However, according to Zayats, the plasmonic device performance can be improved by optimization of its structure.
Semiconductor fabrication company Silios Technologies of Rousset-Peynier, France, hopes to commercialize the group’s technology and other research that comes out of Plasmocom, which is hosting a workshop in June on applied plasmonics.
It may be some time before devices are available commercially, but eventually they might be able to do what cannot be done now, Zayats said. “The ultimate goal is an integrated photonic circuit based on plasmonic excitations capable of performing all operations completely optically.”