A new twist to optical communications

One important step towards such systems has been demonstrated by physicists at the Harvard School of Engineering and Applied Sciences. The team has developed a generalised coupler that converts free space beams to a metal surface by inducing a surface wave known as a surface plasmon polariton (spp). Converting a free space beam "The idea of our work is inspired by the principle of holography," said Patrice Genevet, research associate at Harvard University. "We designed a metallic interface a coupler that converts a free space beam with a complex wavefront into a propagating surface plasmon polariton." An spp is an electromagnetic wave that propagates at the interface between a dielectric – in this case, free space and a metallic film, and is an important element in the emerging field of nano photonics. "These waves propagating at the surface are complicated to couple to; you have to play some tricks," said Genevet. Until now, researchers have had to use simple light excitation, such as plane waves or Gaussian beams, to induce spps. The Harvard researchers have generalised the design of plasmonic couplers using a holographic approach. The couplers are two dimensional holograms that scatter the light of a reference beam to excite a surface wave. "The holographic coupler acts like a bunch of scatterers placed judiciously to phase match the [two] beams," said Genevet. "The coupler transfers energy coherently from one beam the free space beam to the other, the surface wave." To prove the technique, the team designed couplers for complex free space beams, such as vortex beams that rotate in a corkscrew fashion as they travel. Such light beams are characterised by their orbital angular momentum. The holographic couplers are computed by interfering spps with the complex light beams. "Several designs have been proposed to excite spp waves, but no one has designed couplers for beams with such complex waveforms," said Genevet. The team has applied the concept to silicon photodetectors. By patterning the holographic couplers directly on top of standard detectors, the team has demonstrated that photodiodes can not only detect the intensity of light, but also provide information on the number of twists of its wavefront. "Detectors which can directly sort the orbital angular momentum of light are interesting for future optical communication," said Genevet. Current optical systems use wavelength division multiplexing (wdm), in which data is sent down a fibre at different wavelengths. Leading commercial wdm systems can transmit up to 20Tbit through a fibre. Due to the continual growth of internet protocol traffic, researchers are already looking at ways to increase the transmission bandwidth beyond classical wdm systems. One possible approach being explored in R&D labs consists in sending light through multiple fibres and down multiple paths – or modes – per fibre. Bell Labs, Alcatel-Lucent's research arm, has demonstrated the transmission of data at 3.8Tbit/s using several channels and an advanced signal processing technique known as multiple input, multiple output (MIMO). In particular, 40G quadrature phase shift keying (QPSK) signals were sent over a fibre with six paths, or spatial modes. Two polarisation modes and eight wavelengths were also used for each path to achieve the overall capacity of 3.8Tbit/s. The overall transmission only uses 400GHz of spectrum; in comparison, current 100G optical systems can send 800Gbit/s over 400GHz of fibre spectrum.
Orbital angular momentum Another space division multiplexing approach consists of using the orbital angular momentum of light as a means to encode information. The interesting element here is that the number of rotations of the wavefront – referred to as the topological charge of the angular momentum can theoretically take an infinite number of integer values. A free space helical beam generated using a 850nm laser can twist any integer number of times within the 850nm wavelength distance. "So now, instead of encoding information on two bits you either have light or you don't have light you can encode information on the orbital angular momentum of light: you either have a given orbital angular momentum or you don't have light," said Genevet. "Since the orbital angular momentum can cover many integer values, you increase the amount of encoded information dramatically." However, significant practical challenges must first be overcome. In order to detect the differently encoded light beams, they must first be split before being fed to an array of photodetectors. Moreover, each photodetector has to be tailored holographically tuned – to detect one single beam with a specific number of twists. Splitting the signal decreases the amount of light that falls on each detector. Also, the demonstration at Harvard used detectors designed for a free space wavelength of 630nm. In the visible range of wavelengths, metallic interfaces are extremely lossy which results in poor responsivity of the detectors. "The photodetector has a given responsivity," said Genevet. "When we pattern it with the metal film, the efficiency reduces by almost a factor of 1000." But there are ways the losses can be countered. For photodetectors working at the longer 1500nm wavelengths used for telecom, the propagation length of the spp increases significantly due to lower plasmonic losses. Avalanche photodetectors, which have a much higher efficiency, can also be used. Genevet estimates that such changes would counter the factor of 1000 and restore the losses. The team's latest focus is not to develop detectors to enhance optical communication performance, but rather to address more fundamental issues. "For sure, we are interested in data transfer using orbital angular momentum. But we are also curious in investigating fundamental problems and trying to come up with new ways to excite surface waves," said Genevet. The team is exploring how to modify the coupler to detect other types of wavefronts and other beam characteristics such as polarisation. This would add yet another dimension to a detector's measurements alongside intensity and phase rotation. Genevet believes that information processing with orbital angular momentum could increase optical capacity dramatically by encoding information across multiple vortex beams. However, he is not sure about whether optical communications would be the first application of this technology. "One could certainly integrate our devices into optical communication systems, but it is hard to determine at this stage whether or not orbital angular momentum will play the leading role in space division multiplexing technology," he says. In contrast, spps are applicable in areas such as sensing and medicine and these applications might also benefit from the holographic coupling technique.