The trouble with diffraction gratings is that you get more than you ask for---light is sent into several diffraction orders, meaning that the input light is split between several spots. Gratings that have a single output spot are tricky to make, as they contain sharp discontinuities where the phase "wraps" back to zero after a wavelength of retardation, and must modulate the phase by exactly the right amount. Polarisation gratings neatly sidestep this problem, by rotating polarisation rather than directly modulating optical path length. Effectively, a polarisation grating consists of a half wave plate where the fast axis is oriented differently at different points on the surface. For circularly polarised light, this is equivalent to changing the optical path length. However, the orientation of a half wave plate naturally wraps back to zero after rotating the polarisation by 2π. This means that the discontinuities present in most phase gratings are eliminated, and so polarisation gratings can be made very efficient. Usually made by aligning birefringent polymer or liquid crystal molecules with light, these gratings can reach efficiencies of over 95%. The liquid crystal gratings can also be electrically switched from gratings to transparent windows, allowing them to be used as beam steering devices.

Most polarisation gratings are made by aligning the molecules in an interference pattern between two plane waves; this produces a highly efficient grating that deflects light in one direction or the other depending on the handedness of its circular polarisation (much like a Wollaston prism does for linearly polarised light). Li and co-workers have extended this approach by placing a vortex phase plate in one of the writing beams. The resulting hologram still mostly resembles a linear diffraction grating, but with a fork dislocation in the centre. A beam passing through such a grating will now not only have its direction altered, but will pick up an additional phase structure, changing plane waves into spiral wavefronts; it now carries orbital angular momentum.

In the last couple of decades, there has been much interest in light beams carrying orbital angular momentum (OAM); these beams have a spiral phase structure and rays that seem to "twist" around the optical axis, and they have been used to rotate small objects and carry quantum information to give but two examples. Much effort has been devoted to methods of manipulating and measuring the orbital angular momentum of light, using q-plates, static holograms and even computer-controlled spatial light modulators. It is well established that diffractive methods can produce very pure OAM states, usually at the cost of efficiency and/or expensive fabrication. However, these polarisation gratings promise extremely high efficiency (over 90%) at much lower cost than a spatial light modulator.

Crucially, the very pure states generated by a polarisation grating can be switched either electrically (in a few milliseconds) or optically (much faster, rotating the polarisation of the input with a Pockels cell). This could provide a way of modulating the OAM carried by a beam at high speed, which would in turn enable high bandwidth communications by encoding information as the amount of “twist” on the light. In fact, this has been put forward as a way of performing quantum encrypted communications over a free-space link. A communication system able to use multiple spatial states of the light (such as these helically phased modes) could be more secure and/or faster than a technique with only two states (like polarisation-encoding); the improvement is similar to MIMO technology already seen in wireless radio communications. Generating pure OAM states very quickly is one of the main challenges to overcome if these communications systems are to be realised, and the gratings described in this paper bring that one step closer to reality.

---

---

 Add Comment

You must log in to add comments.