Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Research highlights

Non-local distillery

Phys. Rev. Lett. (in the press); preprint at (2008)

One of the most intriguing and mysterious properties of quantum mechanics, non-locality, can be amplified. This is the conclusion that Manuel Forster and colleagues reach in a theoretical study, and they report a protocol for doing so.

Non-locality means that the outcome of measurements on spatially separated quantum systems can be strongly correlated, more strongly than classical physics can explain. Beyond their fundamental interest, non-local correlations serve as a precious resource for quantum-information processing. And for these applications, generally speaking, the stronger the non-locality displayed by a system, the more useful it is.

Forster et al. start with weakly non-local, bipartite systems, and show that stronger non-locality can be obtained when two parties do nothing more than perform local operations on their respective parts. The strength of non-locality is traditionally measured by the extent to which a so-called Bell inequality is violated. But this strength can be changed, as Forster et al. show, and they propose an alternative measure that might be more helpful in determining the usefulness of non-locality.

Storage in a spin

Nature 458, 489–492 (2009)

One of the consequences of Faraday's law, which describes the electromotive force generated by a time-varying magnetic field, is that passing a current through a simple induction coil will inject and store energy in the coil, in the form of a generated magnetic field. Recently, Stewart Barnes and Sadamichi Maekawa argued that Faraday's law should be extended to include non-conservative spin effects resulting from the rotation of spin currents flowing through a ferromagnetic metal. They and colleagues now show experimentally that such effects become evident in the electronic behaviour of magnetic tunnel junctions and, as in a simple coil, result in the storage of energy.

When they subjected a layer of GaAs embedded with zinc-blende-structured MnAs nanoparticles, separated from a continuous layer of MnAs by a thin insulating AlAs layer, to a magnetic field, the tunnelling of spins between the superparamagnetic nanoparticles and the continuous MnAs layer generated an electromotive force. At a field of 1 T, the current–voltage characteristics of the device shifted by 21 mV, and an open-circuit voltage of about 7 mV arose and persisted for longer than 10 min.

Two become three

Phys. Rev. Lett. (in the press); preprint at (2008)

The quantum entanglement of two particles is becoming a routine exercise; the goal now is to entangle ever higher numbers. Toshiyuki Tashima and co-workers have shown how to create a type of three-photon entangled state by the conversion of entangled pairs of photons.

Entangled states having more than two photons fall into distinct classes. It is impossible to convert a state in one class to a state in another by using so-called local operations and classical communication. Therefore, a system that can create any three-photon entangled state must consist of more than three photons. Tashima and colleagues' experiments prove that such a system can be made with just four photons — two entangled pairs.

The authors were able to generate one of the two classes of three-particle entangled states — the W-type states — using polarization-dependent beam splitters and post-selecting specific photons. The creation of the other, Greenberger–Horne–Zeilinger, class of states from two pairs of entangled photons having already been demonstrated, it is now clear that, starting with two entangled pairs and using local operations and classical communication, any three-particle entangled state can be created.


Phys. Rev. D 79, 065004 (2009)

It's feasible, in quantum physics, to have a large negative energy density at a point — and with this comes all sorts of weird possibilities such as traversable wormholes and time machines. Fortunately, to stop things getting out of hand, there are constraints on average or total energy over a volume or line, such as that expressed in the 'quantum interest conjecture': overall, the energy density must be positive; negative energy density somewhere must be more than compensated for by positive energy density elsewhere.

For the example of energy pulses, this means that the amount of negative energy in a pulse is constrained to be more than balanced by a larger positive-energy pulse; the time interval between such pulses is also restricted, according to the conjecture. The net energy of the two pulses, necessarily positive, is the 'quantum interest'.

The quantum interest conjecture has already been proved in (1 + 1) Minkowski space, but now Gabriel Abreu and Matt Visser have taken it into more dimensions. By proving a variant of Simon's theorem for the biharmonic Schrödinger equation, they have reformulated the conjecture for (3 + 1) Minkowski space. In fact, the result can be generalized to any Minkowski space that has an even number of dimensions.

Tense, and relax

Appl. Phys. Lett. 94, 091116 (2009)

Despite its amazing success in electronics, silicon is not ideal for optoelectronics. Its inability to generate light efficiently is well known, but it also suffers from a lack of second-order susceptibility — which is essential for components based on nonlinear optics (all-optical switches, for example). To circumvent this failing, Nick Hon and colleagues propose a silicon-based structure that has an engineered second-order nonlinearity.

The absence of second-order susceptibility is due to the symmetry of the silicon crystal. Hon et al. suggest breaking this symmetry by creating a waveguide of alternating layers of silicon nitride under either tensile or compressive stress — what they call “periodically poled silicon”. To show how it could be put to practical use, the authors performed a numerical study of difference-frequency generation — that is, the interaction of two near-infrared laser beams in the waveguide to produce mid-infrared light. The simulation shows that 'mixing' 1.3-μm and 1.75-μm light in a 2-cm-long structure that has a period of 8 μm would allow the efficient generation of 5.1-μm radiation.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Research highlights. Nature Phys 5, 245 (2009).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing