Cosmics cut off

Phys. Rev. Lett. 101, 061101 (2008)

An upper limit to the energy of cosmic rays has been confirmed by the Pierre Auger Collaboration. Predicted in 1966, the 'GZK cutoff' — named after Kenneth Greisen, Georgiy Zatsepin and Vadim Kuz'min — arises because cosmic-ray protons with energy higher than 6 × 1019 eV should interact with the cosmic microwave background radiation, producing lower-energy protons and pions.

Experimentally, however, the situation was unclear. It was claimed that the Japanese AGASA detector had found several cosmic rays at energies exceeding the GZK cutoff. But, earlier this year, the HiRes collaboration — whose detector comprised two arrays of telescopes, 12.6 km apart, in Utah, USA — published the first five-sigma observation of the GZK cutoff (R. U. Abbasi et al. Phys. Rev. Lett. 100, 101101; 2008).

That result is now confirmed by the Auger data, which correspond to an exposure twice that of the HiRes detector and were collected using an array of 1,600 water-Čerenkov detectors that spans 3,000 km2 of the Pampa Amarilla plain in Argentina.

Rippled entanglement

Phys. Rev. A 78, 012114 (2008)

Credit: © D. CSONTOS

The existence of gravitational waves remains one of the most important, yet unverified, predictions of Einstein's general theory of relativity. Fabrizio Tamburini and colleagues present a fresh approach to the problem, based on entangled photon states.

For example, if Alice and Bob share a secret through quantum-key distribution, encoded in entangled pairs of photons, then eavesdropping Eve cannot intercept the information without altering the entanglement and giving away her presence to Alice and Bob. Tamburini et al. propose that the detection of gravitational waves could work according to the same principle.

Viewed classically, a gravitational wave would alter the local time and reference frame of the two detectors relative to each other and the path-length travelled by the photons. At the quantum level, the give-away would be a change in the cross-correlation statistics, because the graviton–photon interaction would lead to decoherence of the entangled photon pairs. The sensitivity of the scheme, say the authors, would be comparable to the best present-day designs that are based on laser interferometry.

Few moments remain

Nature 454, 976–980 (2008)

The behaviour of electrons in many metals and semiconductors can be rationalized in terms of Fermi-liquid theory, in which electrons behave as independent particles with an effective mass. There are, however, exceptions.

Ncholu Manyala and colleagues have demonstrated the existence of a new class of materials — small-bandgap semiconductors with manganese impurities — that display non-Fermi-liquid behaviour near a metal–insulator transition. Applying a small magnetic field restores the Fermi liquid, suggesting that the so-called under-compensated Kondo effect — in which a localized moment is insufficiently screened by the mobile charges — is the source of the observed behaviour.

The introduction of a manganese atom into iron silicide frees up a positively charged quasiparticle (a hole) and the atom acquires a net moment. Detailed measurements by Manyala et al. show that, even when the semiconductor is cooled down to extremely low temperatures, a few localized moments remain visible to the free charge. This is in contrast to other materials in which the moments are either fully screened by mobile charges or couple fully to each other through antiferromagnetic exchange interactions.

Divide and conquer

Phys. Rev. Lett. 101, 058701 (2008)

Credit: © SPL

It was realized in the late eighteenth century that the relatively benign cowpox virus could be used to immunize humans against smallpox, a lethal disease. However, finding the optimum strategy for the immunization of a population — or, in a modern analogy, for the protection of a computer network — using a minimum number of vaccine doses remains an open problem. Yiping Chen and colleagues have a new strategy, one that they claim requires up to 50% fewer doses than existing schemes.

Mathematically speaking, the problem is how to fragment a network with a minimum number of node removals. Earlier strategies involved knocking out the most central nodes until the network was broken up into disconnected clusters. But this approach led to a broad distribution of cluster sizes, and, therefore, many doses have to be used to isolate relatively small clusters. Chen et al. now demonstrate that fragmentation of the network into clusters of approximately equal size is a more efficient strategy, and provide tools for this 'equal graph partitioning'.

The long and the short of it

Europhys. Lett. 83, 17001 (2008)

The way in which charge carriers scatter is one of the key factors that determines a material's electrical characteristics. Graphene is no exception. But seemingly contradictory results on the influence of adsorbed impurities have confounded attempts to establish a connection between charge scattering and various anomalous aspects of graphene's behaviour — such as its unusually high minimum conductivity.

Maxim Trushin and John Schliemann have constructed a semiclassical microscopic model to describe graphene's electronic behaviour that incorporates both short- and long-range impurity scattering potentials. According to this model, scattering from impurities with short-range potentials — such as nitrogen dioxide — only affects the mobility of carriers close to the Dirac point at low carrier concentrations. Scattering from longer-range impurities — such as potassium — however, affects the carrier mobility over a wider range of energies and concentrations.

The model also suggests that graphene's anomalous minimum conductivity is a result of the chiral nature of its charge carriers, rather than the shape of its bandstructure — which could be proved by future experiments on impurity scattering in bilayer graphene.