Vol. 46 No.2 - Highlights

Neural networks beyond the mean-field paradigm (Vol. 46 No. 2)

Schematic of a hierarchical neural network (center): four stored patterns (top), with two retrieved (bottom).

The seminal paper "Neural Networks and physical systems with emergent collective computational abilities" by Hopfield (1982) and its statistical mechanical treatment by Amit, Gutfreund and Sompolinsky (1985) still play as "harmonic oscillators" in Artificial Intelligence: crucially, in their picture, "associative memory" emerges as a collective feature of neurons. Due to mathematical constraints, this paradigmatic formalisation relies on the so-called "mean-field" approximation: each neuron interacts with all others in the network, regardless of their reciprocal distance. As a (non-obvious) consequence, the network performs “serial processing'': it is able to retrieve one pattern of information per time. Here we show a way, based on a hierarchical underlying topology (see figure), to overcome mean-field limitations, thus accounting for neuronal distance in the network (this also allows for dilution as neurons too far away do not interact). Remarkably, simply introducing a metric (that is a biological must) enables the network to spontaneously switch from serial processing to parallel processing: it can retrieve several patterns of information simultaneously. These emergent multitasking features characterize a novel generation of neural networks, which better capture real brain behavior.

E. Agliari, A. Barra, A. Galluzzi, F. Guerra, D. Tantari and F. Tavani, “Metastable states in the hierarchical Dyson model drive parallel processing in the hierarchical Hopfield network”, J. Phys. A: Math. Theor. 48, 015001 (2015)

Plasma density evolution in a microwave pulse compressor (Vol. 46 No. 2)

Plasma density vs. time after the laser triggering for different microwave output.

Microwave plasma discharges have been widely investigated for many years; there is, however, a subject that is insufficiently studied. It is the plasma formation at the initial – nanosecond time-scale – stage of the high-pressure discharge in a resonant cavity and its interrelation with the process of microwave energy release from the cavity that goes out of resonance during the plasma generation. This subject directly concerns the operation of microwave compressors using commercial magnetrons and klystrons for short-pulse high-power microwave generation. In this work, for the first time, spectroscopic measurements were performed to investigate nanosecond dynamics of the plasma density in the S-band compressor with laser triggering. For pressurized helium filling the compressor cavity and switch, the plasma density was evaluated from the shapes of the 3888.65 Å and 4471.5 Å He I spectral lines. The measured evolution of the density was found to correlate with the peak power of the compressor output pulse and efficiency of the stored microwave energy extraction. With increasing microwave output, the plasma appears earlier in time after the laser beam enters the system, the plasma density rises more steeply, and it reaches higher values.

L. Beilin, A. Shlapakovski, M. Donskoy, T. Queller and Ya. E. Krasik, “Plasma density temporal evolution in a high-power microwave pulse compressor switch”, EPL 109, 25001 (2015).

Finite temperature entanglement in many body systems (Vol. 46 No. 2)

Entanglement negativity at finite temperature between an interval A (red) and the remainder (green) obtained through particular correlation functions on an infinite cylinder whose radius is proportional to the inverse temperature.

Entanglement is a key feature of quantum mechanics setting it apart from the classical world. In the last decade, entanglement also became a practical tool to characterise the various phases of matter of many-body quantum systems in a pure state, in particular in connection with topological and critical phases. However, the quantification of entanglement for a bipartite many-body system in a mixed state, such as at finite temperature, is a harder task. Various measures of entanglement for mixed states have been introduced and the most practical one is the so-called negativity. Focusing on one-dimensional many-body systems at criticality, for a bipartition of the system into a finite interval and its remainder (see figure), we find an expression for the negativity at finite temperature, which turns out to depend only on the ratio between the temperature and the length of the interval. This universal function encodes the full operator content of the theory.

P. Calabrese, J. Cardy and E. Tonni,, “Finite temperature entanglement negativity in conformal field theory”, J. Phys. A: Math. Theor. 48, 015006 (2015)

The power of light-matter coupling (Vol. 46 No. 2)

Illustration of molecules coupled to the fundamental optical mode of a 145nm thick Fabry-Perot cavity. It features a typical example of the absorption spectrum of uncoupled (red line) and coupled (dark line) molecules.
Credit: A. Canaguier-Durand et al.

A theoretical study shows that strong ties between light and organic matter at the nanoscale open the door to modifying these coupled systems’ optical, electronic or chemical properties.

Light and matter can be so strongly linked that their characteristics become indistinguishable. These light-matter couplings are referred to as polaritons. Their energy oscillates continuously between both systems, giving rise to attractive new physical phenomena. Now, the authors have explained why such polaritons can remain for an unusual long time at the lowest energy levels, in such a way that alters the microscopic and macroscopic characteristics of their constituting matter. These new results are in agreement with experimental results. These findings thus pave the way for optical, electronic and chemical applications.

A. Canaguier-Durand, C. Genet, A. Lambrecht, T. W. Ebbesen, and S. Reynaud, “Non-Markovian polariton dynamics in organic strong coupling”, Eur. Phys. J. D 69, 24 (2015)

Analogue quantum computers: still wishful thinking? (Vol. 46 No. 2)

Quantum particle with a wave-function delocalised over the entire potential energy landscape can "feel" a deep isolated potential minimum (hole) on a flat potential energy landscape (table) even before "falling" (completely localising) into it.

Many challenges lie ahead before quantum annealing, the analogue version of quantum computation, contributes to solve combinatorial optimisation problems

Traditional computational tools are simply not powerful enough to solve some complex optimisation problems, like, for example, protein folding. Quantum annealing, a potentially successful implementation of analogue quantum computing, would bring about an ultra-performant computational method. A series of reviews guest-edited by the authors, focuses on the state of the art and challenges in quantum annealing. This approach, if proven viable, could greatly boost the capabilities of large-scale simulations and revolutionise several research fields, from biology to economics, medicine and material science.

A company called D-Wave has been commercialising what it claims are quantum annealers, since 2011. There have been speculations from the science community as to whether the D-Wave technology actually delivers quantum annealing. “The reviews of our latest issue show that the performances of the D-Wave machines as quantum computers, while noteworthy, have remained essentially inconclusive,” explain the authors.

S. Suzuki and A. Das, “Quo Vadis quantum annealing?, Eur. Phys. J. Special Topics 224, 5 (2015)

A. Das and S. Suzuki,, “Debate and discussion: Quo Vadis quantum annealing?”, Eur. Phys. J. Special Topics 224, 205 (2015)

Picosecond-range control over information processing (Vol. 46 No. 2)

Energy levels of the parabolic QD versus the strength of the Rashba SOC Credit: J. A. Budagosky et al.

Optical manipulation is key to reaching the necessary speed to control the furtive underlying physical mechanism used in quantum information processing.

Quantum computing will, one day, bring quicker information processing. One of the keys to such speed is being able to control the short-lived physical phenomenon holding quantum information, also known as quantum bits (qubits). A new study presents a novel optical manipulation technique to control one possible kind of qubit—represented, in this case, by polarised electron spins—exposed to an ultra-short pulsed laser in the picosecond-range. The authors have tested this novel optics approach using a quantum dot—nanoscopic artificial structures with a small number of electrons—in this study.

They used optical manipulation relying on very high-frequency—terahertz—laser pulses to induce a 180⁰ rotation of the polarisation of the spin of a single electron confined in a semiconductor quantum dot. They then used a set of mathematical tools to define the most effective manipulation technique.

J. A. Budagosky Marcilla and A. Castro, , “Ultrafast single electron spin manipulation in 2D semiconductor quantum dots with optimally controlled time-dependent electric fields through spin-orbit coupling”, Eur. Phys. J. B 88, 15 (2015)

Ultrafast laser for crafting ever thinner solar cells (Vol. 46 No. 2)

Interaction between ultrashort pulse laser and a target.

Solar-cell efficiency depends on how thin it can be manufactured. Now, a new model exploits femtosecond laser sources to get higher efficiency at lower cost.

The race for ever more efficient and cheaper solar cells tests the limits of manufacturing. To achieve this, photovoltaic solar cells need to become thinner and are made of more complex inner structures. Now, the authors have investigated and expanded a model elucidating the dominant physical processes when ultra-fast lasers are used in manufacturing solar cells to these specifications.

The authors rely on ultra-fast lasers, to develop a process called ablation, used to allow the formation of the metal contacts. It involves selectively removing the upper dielectric layer of the photovoltaic cell material without damaging the semiconductor beneath. Compared to previous methods, it offers many advantages—it reduces heat damage while improving the precision, energy efficiency and speed of the process.

A. Gurizzan and P. Villoresi, “Ablation model for semiconductors and dielectrics under ultrafast laser pulses for solar cells micromachining”, Eur. Phys. J. Plus 130, 16 (2015)

Atoms crystallised by light for precision measurement (Vol. 46 No. 2)

Trajectories and intensities of two beam splitters. Credit: D. Holzmann et al.

A new study focuses on the collective dynamics of illuminated atoms coupled to photons travelling in a nanofibre.

Theoretical physicists have uncovered the existence of self-organised dynamics of atoms, bound by light into a crystal, with long range atom-atom interactions. These findings were recently obtained by the authors. This approach could, among others, help to better understand the process of crystallisation in new materials, and help implement efficient photon storage and precision measurements.

Their study focuses on atoms trapped in the leaked light very close to a tapered optical nano-fibre. Such fibres are too thin to confine all the light within. These atoms are exposed to a transverse laser beam, whose light becomes partially redirected, or scattered, into the nanofibre at each atom’s position, before propagating along the chain of atoms. It thus mediates a strong effective atom-atom interaction. This approach yields a stable chain of atoms, bound by light, forming a crystal.

D. Holzmann, M. Sonnleitner and H. Ritsch, “Self-ordering and collective dynamics of transversely illuminated point-scatterers in a 1D trap”, Eur. Phys. J. D 68, 352 (2014)

New experimental approach for near–edge ultrafast EUV absorption spectroscopy (Vol. 46 No. 2)

Ultrafast absorption spectroscopy at the L(2,3)-edge of silicon

Time resolved pump-probe spectroscopy in the extreme ultraviolet (EUV) spectral range offers the opportunity to study properties and structure of the electron subsystem in condensed materials under non-equilibrium conditions rapidly changing in the sub-ps time scale. New frontiers studies can be accomplished thanks to the availability of new generation sources such as free electron lasers (FEL). The breakthrough research interest in the EUV radiation–matter interactions requires the development of pivotal optical elements able to manipulate short wavelength beams. Conventional single layers coated mirrors provide negligible reflectance in the extreme ultraviolet spectral range, therefore knowledge coming from other disciplines is required to overcome such technological limits. The development of the multilayer coated mirrors has been intensively driven by the microelectronics industry in view of their application in EUV lithographic apparatus. The same technology has been used to develop a novel broadband multilayer coated mirror conceived specifically for near-edge ultrafast absorption spectroscopy. Such an optical element has been inserted in the EIS-TIMEX end-station at FERMI@ELETTRA FEL. The design of the device has been optimised in order to manipulate the FEL pulses preserving their temporal and wavefront properties in the wavelength range required by the ultrafast absorption spectroscopy at L(2,3)-edge of silicon (see figure).

A. J. Corso, P. Zuppella, E. Principi, E. Giangrisostomi, F. Bencivenga, A. Gessini, S. Zuccon, C. Masciovecchio, A. Giglia, S. Nannarone and M. G. Pelizzo, "Broadband multilayer optics for ultrafast EUV absorption spectroscopy with free electron laser radiation", J. Opt. 17, 025505 (2015)

Efficient Lattice-Boltzmann simulation techniques for nonlinear thermoacoustic engines (Vol. 46 No. 2)

Thermoacoustics is the physics of the interaction of thermal and acoustic fields. The nonlinear acoustic effect and low Mach number compressible flow in thermoacoustic engines make the theoretical analysis of such systems extremely complicated. A new study investigates the nonlinear self-excited thermoacoustic onset in a Rijke tube via the lattice Boltzmann method (LBM), which simulates the fluid flow by tracking the evolution of particles and obtains flow stream and heat transfer patterns from the kinetic level. The adopted LBM model, which was developed by the authors, convincingly simulates the Navier-Stokes-Fourier equations, treating accurately the nonlinear process of wave excitation of coupled fields and providing reliable estimates for pressure, density, velocity and temperature in such a finite geometry.

A nonlinear self-excited standing wave in the Rijke tube is observed from simulations. Agreement is obtained with theoretical predictions when they exist. Instantaneous velocity fields and temperature fields are discussed. The maximal Mach number in the Rijke tube is about 0.035, indicating that the air flow under the limit cycle is the low Mach number compressible flow.

Y. Wang, D.-K. Sun, Y.-L. He, and W.-Q. Tao, “Lattice Boltzmann study on thermoacoustic onset in a Rijke tube”, Eur. Phys. J. Plus, 130, 9 (2015)

Mechanisms of two-proton emission seen in three-body correlations (Vol. 46 No. 2)

Sequential two-proton decay of the 16Ne Er=7.57 MeV state. The fractional energy distribution (left) gives resonance energy in 15F while the angular distribution (right) determines Iπ of the initial state.

Hitherto three-body correlations between decay products of nuclear resonances, unstable to the emission of two neutrons have been a very effective tool in the analysis of GSI-experiments on 5H, 10He, 13Li, and 14Be. Here the first report is given about the mechanisms for two-proton emission from states in 16Ne, representing the presently most complete study of this nucleus. One-neutron knockout from 17Ne populated the 16Ne(g.s.) (Er=1.39 MeV, Γ=0.08 MeV) above the 14O+p+p threshold, and resonances at Er=3.22 MeV and 7.57 MeV. The decay mechanisms were revealed analysing three-body energy correlations in the 14O+p+p system. It was found that the 16Ne(g.s.) undergoes a democratic three-body decay. In contrast to this, the 16Ne(21+) state emits protons through the 15F(g.s.) sequentially. The decay of 7.57 MeV state is well-described assuming emission of a proton from the d5/2 shell to 15F(5/2+), which decays by d5/2 proton emission to 14O(g.s.). By using R-matrix analysis and mirror symmetry this state was unambiguously identified as the third 2+ state in 16Ne.

J. Marganiec et al. (+ 58 co-authors), “Studies of continuum states in 16Ne using three-body correlation techniques”, Eur. Phys. J. A 51, 9 (2015)

Optical manipulation of particles of all shapes and sizes (Vol. 46 No. 2)

Ellipsoidal particle levitating at the optical beam waist, with its long axis aligned to the beam axis. Credit: B. M. Mihiretie et al.

A new study of how particles may respond to the mechanical effects of light helps improve optical manipulation and remote sensing of non-spherical particles.

Manipulation of small objects by light has gained in popularity in the past few years. Now, scientists have performed the first systematic analysis of the behaviour of ellipsoidal particles manipulated by laser beams. The work shows that such particles are constantly moving in and out of the reach of an optical beam, creating oscillations. These findings have been obtained by the authors. This work could help understand the unusual behaviour of rod-like particles manipulated by optical tweezers. Ultimately, the theoretical part of this work could contribute numerical models of how complicated shapes and large sizes scatter laser light. Numerous applications exist in fluid engineering and remote sensing methods.

B. M. Mihiretie, P.Snabre, J.-C. Loudet and B. Pouligny, "Optically driven oscillations of ellipsoid particles. Part I: Experimental observations", Eur. Phys. J. E 37, 124 (2014)

J.C. Loudet, B. M. Mihiretie and B. Pouligny, "Optically driven oscillations of ellipsoidal particles. Part II: Ray-Optics calculations", Eur. Phys. J. E 37, 125 (2014)