Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://qi.phys.msu.ru/papers/2010-nphoton-4-585.pdf Äàòà èçìåíåíèÿ: Wed Jun 29 12:09:41 2011 Äàòà èíäåêñèðîâàíèÿ: Thu Feb 27 20:32:40 2014 Êîäèðîâêà:

news & views

BIOPHOTONICS

Fluidic camera aids surgery
Medical staff may soon benefit from a much clearer view during key-hole surgery (laparoscopy), thanks to the invention of a miniature fluidic zoomable camera that provides unprecedented image quality for surgical applications (pictured, left). The camera, developed by Frank Tsai, Yu-Hwa Lo and co-workers from the University of California at San Diego (UCSD) in the USA, was presented at the Conference on Lasers and Electro-Optics in San Jose in May 2010. It combines LED illumination with a tiny 4â zoom lens (17 mm long with a 4 mm clear aperture) made from a fluid-filled membrane lens assembly. The lens assembly is integrated with a 2-megapixel CMOS imaging chip and is packaged in a protective housing containing the necessary electronics and actuators. An electrochemical liquid pump is used to change the shape of the lens membrane surface between concave and convex, allowing both wide-angle and telephoto/ macroscopic imaging to be achieved. "In our fluidic lens autofocusing system, the device can operate either as a regular camera, a macro-mode camera for resolving 0.1-mm-sized features, or, with some assistance from image processing, a microscope for resolving features as small as 2­3 m," explained Lo, who leads the research group. "The image quality is currently between SVGA (800 â 600 pixels) and HD (1280 â 720 pixels). With additional lenses to correct for aberrations and colour dispersion, HD-quality imaging can be assured."

According to Lo, the approach has many benefits over the laparoscopic cameras traditionally used in key-hole surgery. Conventional laparoscopes, which consist of a simple rod lens and a fibre imaging bundle, usually require very high powers of illumination (~20,000 lux) for effective imaging, and have a fixed, limited field-ofview together with a fixed focus (that is, no zooming or autofocusing capabilities). In contrast, the fluidic camera from Lo and colleagues has a very large focal length (diopter) tunability and can operate in light levels as low as 300 lux. "We have demonstrated that the fluidic lens imaging system can achieve a tuning range of over 200 diopters, which is more than ten times greater than any fixed lens system of the same size," explained Lo. "The fluidic lens has an intrinsic response of a few milliseconds for autofocusing

and less than one second for 4â optical zoom." Working with clinicians at UCSD and the Hadassah­Hebrew University Medical Center in Jerusalem, the team have already used the camera during gall bladder removal surgery on pigs (pictured, right). It is hoped that the camera will prove useful for many forms of surgery and medical imaging, including for cancer surgery, retinal imaging and opthalomology. Rhevision Technology in California has recently licensed the technology and is now in the process of commercializing it. By using conventional optics in combination with the fluidic lens to correct for aberrations, the firm has shown that it is possible to achieve an image quality as good as SLR cameras.
OLIVER GRAYDON

QUANTUM OPTICS

The quest for higher dimensionality
Using a clever design of polarization optic, Italian researchers have successfully created four-level `ququart' quantum states using the polarization and orbital angular momentum of single photons. This approach may help to realize more effective forms of quantum communication.

Stanislav Straupe and Sergei Kulik

E

normous progress has been made in the field of quantum communication over the past two decades. What was merely a collection of beautiful ideas in the late 1980s has now developed into

an interdisciplinar y area of science with emerging technological applications. In quantum communication, information is usually encoded in qubits -- quantum systems with two orthogonal states that

are used to represent a binar y digital bit of information (one or zero). However, there is increasing interest in exploiting protocols that use systems of higher dimensionality, which have a greater number of logic levels
585

NATURE PHOTONICS | VOL 4 | SEPTEMBER 2010 | www.nature.com/naturephotonics

© 2010 Macmillan Publishers Limited. All rights reserved

Y. H. LO


news & views
than binar y systems. At a generic level, multilevel quantum systems are called qudits, with specific names given according to the number of levels; for example, qutrits for three levels and ququarts for four levels. Qudits are exciting because of the advantages they offer for quantum information processing. For example, quantum key distribution -- one of the most developed and commercially relevant applications of quantum information science -- is more secure and noise resilient with qudit-based encoding. Furthermore, qudits may also aid in the realization of more efficient quantum gates. Optical qudits can be realized using the various degrees of freedom of a single photon, such as polarization, momentum, frequency, timing or even the spatial shape of a single-photon beam. In recent years it has become popular to use states that have definite values of orbital angular momentum (OAM) as qudits1. However, OAM states are generally difficult to manipulate, as they require complex wavefront transformations usually implemented using holographic techniques. These techniques are rather slow, and thus limit the bit rates that can be achieved. This is a serious problem, particularly because efficient quantum communication requires high bit rates. Fabio Sciarrino and co-workers2 now introduce a convenient method of transferring information from polarization to OAM and vice versa, thus providing a convenient and elegant way for fast polarization controllers to manipulate OAM qudits. A natural basis for OAM states is a set of Laguerre­Gaussian modes, which are eigenmodes of the wave equation under the paraxial approximation. The wavefronts of such modes are composed of m intertwined helical surfaces with a phase singularity in the centre of the beam. Each photon in the beam has an OAM of value m. If the beam is circularly polarized, it also has a spin angular momentum of ± per photon. Usually, when the beam propagates in a homogeneous and isotropic media, these two kinds of angular momentum are separately conser ved. However, in a specially designed medium that is both inhomogeneous and spatially anisotropic, spin and orbital angular momentum may be transferred between each other, as shown by the authors several years ago3. This occurs in particular if a circularly polarized beam passes through a slab of birefringent material with a fixed phase shift of between the polarization components (as in a regular half-wave plate), but with the orientation of its fast axis (optical axis)
586

Optical

axis

Figure 1 | The q-plate is a half-wave plate with an optical axis that has a spatially varying orientation, as indicated by the dark blue concentric circles (and the red vector for one example location). This inhomogeneous birefringent medium is used to convert spin (polarization) to orbital angular momentum, and may prove useful in improving the security of quantum communications.

var ying according to the location of the beam on the plate. A `q-plate' is a plate in which the orientation of the optical axis obeys the relation (r,) = q + 0, where r and are the radial and angular polar coordinates in the plane of the plate, and q and 0 are constants. An example of a q-plate with q = 1 is shown in Fig. 1. In this particular case, if a left (L)-circularly polarized beam passes through the plate, it becomes right (R)-circularly polarized and also acquires an OAM of 2 per photon. The authors made such a birefringent plate using a nematic liquid-cr ystal layer of an appropriate thickness placed between two polyamide-covered glass plates. Circular orientation of the liquid cr ystals was achieved by scratching the surface with a rotating piece of fabric. When a q-plate is used with single photons (that is, a non-classical state of light), its unique property of transferring information from polarization to OAM (and vice versa) may be used to generate entanglement between these two degrees of freedom. In such an entangled state, the wavefunction describing the joint state cannot be factorized into separate spin and OAM components. Imagine a horizontally (H)-polarized photon passing through a q-plate. Because an H-polarized state may be regarded as an equally weighted superposition of R- and L-circularly polarized states, the output of the system turns out to be a Bell state -- a state exhibiting maximum entanglement 4 -- such that: |H|0
o

as `ququarts', which are encoded in both the OAM and polarization of a single photon. Among the various states in four-dimensional Hilbert space, a specific set of bases that are `mutually unbiased' is of particular importance. These form a complete set of five orthogonal bases in such a way that a state |i belonging to any basis has equal projections onto the states of all the other bases |j, giving i|j = 1/4. Thus, all the other states look the same when obser ved from any basis -- a property that is crucial for the security of quantum key distribution protocols. It can be shown that there are exactly five mutually unbiased bases in a four-dimensional Hilbert space, and that a quantum key distribution protocol becomes more secure as more of these bases are used. Until now, the problem has been that two of these bases consist of entangled states, which are difficult to manipulate -- this is where the approach developed by Sciarrino et al. becomes particularly helpful. The authors have succeeded in preparing and analysing states belonging to all five mutually unbiased bases. At the preparation stage, a qubit encoded in photon polarization is transferred to OAM with a q-plate, after which a second qubit is re-encoded in polarization to obtain a hybrid ququart state. For detection, the polarization qubit is first read out with a standard polarization analyser, following which the OAM qubit is transferred to a polarization qubit and then analysed in the same way. The scheme is attractive because it combines the rich capabilities of encoding information in OAM space with the ability to use simple and relatively fast polarization controllers as a manipulation tool. However, applications will probably be limited to free-space quantum communication as Laguerre­ Gaussian modes are not the eigenmodes for ordinar y optical fibres, making OAM encoding inapplicable for fibrebased communication.
Stanislav Straupe and Sergei Kulik are at the Department of Quantum Electronics, Faculty of Physics, M.V.Lomonosov Moscow State University, Leninskie gory, 1, str.2, Moscow 119992, GSP-2, Russia. e-mail: straups@yandex.ru References
1. Molina-Terriza, G., Torres, J. P & Torner, L. Nature Phys. 3, 305­310 (2007). 2. Nagali, E., Sansoni, L., Marrucci, L., Santamato, E. & Sciarrino, F. Phys. Rev. A 81, 052317 (2010). 3. Marrucci, L., Manzo, C. & Paparo, D. Phys. Rev. Lett. 96, 163905 (2006). 4. Nagali, E. et al. Phys. Rev. Lett. 103, 013601 (2009).

(|L|-2o + |R|+2o)/2

where the and o subscripts denote polarization and OAM states, respectively. Sciarrino et al. use this approach to generate four-dimensional qudits known

NATURE PHOTONICS | VOL 4 | SEPTEMBER 2010 | www.nature.com/naturephotonics

© 2010 Macmillan Publishers Limited. All rights reserved