Journalof Physics: Condensed Matter - UCLchianti.geol.ucl.ac.uk/~dario/pubblicazioni/jop_top_papers04.pdf · 2 CONDENSED MATTER: TOP PAPERS 2004 SHOWCASE Magnetism in carbon nanotubes

Top Papers 2004 Showcase highlightsthe leading and most frequentlydownloaded papers, letters andtopical reviews to Journal of Physics:Condensed Matter, arguably theworld’s most authoritative sourceof topical information forcondensed matter physicists andmaterials and surface scientists.Journal of Physics: Condensed Mattercovers experimental andtheoretical studies of thestructural, thermal, mechanical,electrical, magnetic and opticalproperties of condensed matterand is published 50 times per yearby Institute of Physics Publishing.

We welcome papers andsubmissions from authors whohave research interests in theseareas. Authors can benefit fromrapid receipt-to-publication times,a rising impact factor and a largeand growing world widereadership both in print and on-line. On average papers aredownloaded on-line over 100times via our award winningElectronic Journals service.For more information on thesubmission process and to viewthe latest published papers visitwww.iop.org/journals/jpcm

Condensed MatterJournal of Physics:

TOP PAPERS 2004

CO N D E N S E D MAT T E R: TOP PAPERS 2004 SHOWCASE 1

PU B L I S H E D B Y IN S T I T U T E O F PH YS I C S PU B L I S H I N G

S H O W C A S E

Organic and molecular magnets 10S J Blundell and F L Pratt

Coarse grain simulation of soft matter 10Steve O Nielsen, Carlos F Lopez, Goundla Srinivas and Michael L Klein

Simple models of protein folding 10R A Broglia, G Tiana and D Provasi

Self-assembled monolayers 11Frank Schreiber

Quantum cascade structures 11Michael Woerner, Klaus Reimann and Thomas Elsaesser

Slow and fast light in solids 11M S Bigelow, N N Lepeshkin and R W Boyd

Read-out of single spins 12F Jelezko and J Wrachtrup

Superconducting nanostructures 12J G Rodrigo, H Suderow, S Vieira, E Bascones and F Guinea

T O P I C A L R E V I E W S A N D S P E C I A L I S S U E P A P E R S

P A P E R S

Inside you will find synopses of all the top papers and reviews listedabove, together with details of where to find the full articles ➔➔➔➔

Magnetism in carbon nanotubes 2O Céspedes, M S Ferreira, S Sanvito, M Kociak and J M D Coey

Superconductivity and electron correlation 2in a ternary oxideS Yonezawa, Y Muraoka, Y Matsushita and Z Hiroi

Network formation in soft-solid composites 2Doris Vollmer, Gerald Hinze, Wilson C K Poon,Julie Cleaver and Michael E Cates

Unexpected atom exchange in surface diffusion 3Duncan J Harris, Mikhail Yu Lavrentiev,John H Harding, Neil L Allan and John A Purton

Surface recrystallization by electrons 3Tetsuya Narushima, Masahiro Kitajima and Kazushi Miki

Non-linear resonse to ultra-short laser pulses 3C Timm and K H Bennemann

Simulation of highly disordered solids 4I T Todorov, N L Allan, M Yu Lavrentiev,C L Freeman, C E Mohn and J A Purton

Glass transition of polymer films 4H Liem, J Cabanillas-Gonzalez, P Etchegoin and D D C Brandley

Making a diode from isomers 4Anthony N Caruso, Ravi B Billa, Snjezana Balaz,Jennifer I Brand and P A Dowben

Anisotropy in magnetic semiconductors 5Tomasz Dietl

Energy landscape of glassy polystyrene 5V Bercu, M Martinelli, C A Massa, L A Pardi and D Leporini

Charged critical fluctuations in an oxide superconductor 5T Schneider, R Khasanov, K Konder, E Pomjakushina,R Bruetsch and H Keller

Unusual magnetic structures in Gd2Ti2O7 6J R Stewart, G Ehlers, A S Wills, S T Bramwell and J S Gardner

Ferroelectric nano-capacitors 6M M Saad, P Baxter, R M Bowman, J M Gregg,F D Morrison and J F Scott

Heat dissipation in nanostructures 6Andrew P Horsfield, D R Bowler and A J Fisher

Multiple spin-relazation in CMR materials 7R H Heffner, D E MacLaughlin, G J Nieuwenhuys and J E Sonier

Fatigue crack closure micromechanisms 7K H Khor, J-Y Buffière, W Ludwig, H Toda,H S Ubhi, P J Gregson and I Sinclair

Modelling charged colloidal suspensions 7Hiroya Kodama, Kimiya Takeshita, Takeaki Araki and Hajime Tanaka

Wall tensions of colloid-polymer mixtures 8Paul P F Wessels, Matthias Schmidt and Hartmut Löwen

Training effect in CMR manganites 8D Zhu, V Hardy, A Maignan and B Raveau

Time and length scaling in spin glass dynamics 8L Berthier and A P Young

Novel quantum Monte Carlo technique with linear scaling 9D Alfé and M J Gillian

Watching diamond form 9Wataru Utsumi, Taku Okada, Takashi Taniguchi,Ken-ichi Funakoshi, Takumi Kikegawa,Nozomu Hamaya and Osamu Simomura

Phase transitions in quasi-2D structures 9D N McIlroy, S Moore, Daqing Zhang, J Wharton,B Kempton, R Littleton, M Wilson, T M Tritt and C G Olson

Water transport across a capped nanotube embedded in a lipid bilayer.A snap-

shot of water (blue) diffusing across the HBTC tube embedded in a CG-

DMPC bilayer. The hydrophobic part of the tube is depicted in purple, and

the hydrophilic caps in red. The lipid tails (yellow), phosphate unit (purple),

choline unit (red), glycerol and ester groups (dark blue) and semitransparent.)

(See page 10)

2 CO N D E N S E D MAT T E R: TOP PAPERS 2004 SHOWCASE

Magnetism in carbon nanotubesContact induced magnetism in carbonnanotubesO Céspedes, M S Ferreira, S Sanvito,M Kociak and J M D CoeyJ. Phys.: Condens. Matter 16 No 10 L155-L161

Diamond and graphite have many outstandingproperties, but magnetism is not expected to be veryevident in such a light element. There have beenreports of weak ferromagnetism in graphite andother forms of carbon due to a magnetic proximityeffect. Michael Coey’s group at Trinity CollegeDublin set out to find direct experimental evidence of contact-induced magnetism, and gain someunderstanding of its origin. Their idea was to place a carbon nanotube in contact with a ferromagnet and measure the spin transfer associated with thealignment of their chemical potentials. The problemof detecting the tiny spin transfer against the hugebackground magnetic moment of the ferromagnetwas resolved by taking a smooth ferromagnetic thinfilm as a substrate and looking for a stray field aroundthe nanotube. Uniformly magnetized thin filmscreate no stray field, whatever their direction ofmagnetization, so any observed stray field must arisefrom the tube.

Tubes were placed on the different substrates, andtopographic and magnetic images were recorded byatomic and magnetic force microscopy(AFM/MFM). Magnetic contrast was observed forcarbon nanotubes placed on cobalt or magnetitesubstrates, but was absent on silicon, copper or goldsubstrates.

The figure shows some images of a carbonnanotube on Fe3O4 taken at different scan heights.The magnetic images in figures (a) and (b) show littlecontrast to the magnetite substrate, and clear bipolarcontrast to the nanotube, as expected if it were asingle domain magnetized along its length. Theseimages suggest that the tube is aligned with thedirection of magnetization of a single, in-planedomain in the substrate. Shape anisotropy of the filmensures that its magnetization lies in-plane, and anyweak stray field comes from surface irregularity,ripple domains or Bloch walls. The images in figures(c) and (d) are of the same tube in a slightly differentposition.

These studies give direct evidence for contact-induced magnetism due to spin-polarized chargetransfer at a contact between a ferromagnet and acarbon nanotube with a spin transfer of the order of0.1µB per contact carbon atom, and an inducedmagnetization of the order of 1 kA m- 1 in multiwallednanotubes.

This observation of contact-induced magnetismopens a new avenue for implementing spin-electronics at the molecular level, where the currentleads can be separated from the electrodes producingspin polarization.

Superconductivity and ElectronCorrelation in a ternary oxideSuperconductivity in a pyrochlore-related oxide KOs2O6S Yonezawa, Y Muraoka, Y Matsushita and Z HiroiJ. Phys.: Condens. Matter 16 No 3 L9-L12

Pyrochlore oxides have a general chemical formulaA2B2O7 or A2B2O6O�, where A is a larger cation andB is a smaller transition metal (TM) cation. Mostpyrochlore oxides containing 5d TM elements suchas Re, Os and Ir are bad metals. Recently,superconductivity was found for the first time in apyrochlore oxide in Cd2Re2O7 at Tc = 1 K. A relatedpyrochlore oxide Cd2Re2O7 undergoes a metal-insulator (MI) transition at 225 K. The majordifference between them seems to be the number ofd electrons on the B-site cations: Re5+ being 5d2 andOs5+ 5d3.

Z Hiroi and co-workers at The University of Tokyoreport the discovery of superconductivity in a newternary phase KOs2O. Polycrystalline samples wereprepared from KO2 and OsO2 powders. Resistivitymeasurements showed a superconducting transitionwith onset temperature 9.9 K and zero resistivitybelow 9.0 K. When a magnetic field was applied, thetransition curve shifted to lower temperaturessystematically. The resistivity above the transitionshows a peculiar temperature dependence, which isfar from that of a conventional metal. Measurementson a single crystal instead of a polycrystalline pelletare needed to clarify this point. The sample alsoshowed a Meissner effect below 9.6 K. Thesuperconducting volume fraction estimated at 2 Kfrom the zero-field cooling experiment is about 80%,which is large enough to constitute bulksuperconductivity.

The authors believe that an interesting physicalprocess is involved in this compound on the basis ofelectron correlations near the MI transition as well as frustration on the pyrochlore lattice.

See also a recent review by P W Anderson et al.

The physics behind high-temperature superconducting cuprates: the ‘plainvanilla’ version of RVBP W Anderson, P A Lee, M Randeria,T M Rice, N Trivedi and F C ZhangJ. Phys.: Condens. Matter 16 No 24 R755-R769

In this review, Anderson and co-workers revisit theRVB theory using the ‘renormalized mean fieldtheory’. They observe that it is able to explain theexistence of the pseudogap, properties of nodalquasiparticles and approximate spin-chargeseparation, the latter leading to largerenormalizations of the Drude weight and superfluiddensity. Finally, they remark that superexchange, andnot phonons, is responsible for d-wavesuperconductivity in the cuprates.

Network formation in soft-solid compositesThe origin of network formation in colloid-liquid crystal compositesDoris Vollmer, Gerald Hinze, Wilson C K Poon,Julie Cleaver and Michael E CatesJ. Phys.: Condens. Matter 16 L227-L233

Network formation in soft solids affects the way theyrespond to stress and many other properties. In mixedliquid crystal / colloid systems, the tendency of liquidcrystal molecules to align will affect phase structures.What has been less obvious is the role of small alkaneimpurity molecules, which prove to “tune” thekinetics of phase separation.

When liquid crystals are mixed with colloids, theparticles disturb the long-range orientational orderthe liquid crystal molecules can adopt. Consequently,the orientational elasticity of the nematic liquidcrystals expels the colloids, so typically the suspensionphase separates macroscopically into an almost pureliquid crystalline phase coexisting with a phase rich incolloids. However, long-lived non-macrophase-separatedmorphologies have also been observed.

Wilson Poon and co-workers at Edinburgh andMainz have investigated network formation bycomparing detailed rheological, nuclear magneticresonance (NMR) and calorimetric measurements.They studied suspensions of 4-n-pentyl-4�-cyanobiphenyl (5CB) and polymethylmethacrylateparticles (PMMA) with nominal radii of 120, 370 and780 nm, sterically stabilized by chemically graftedpoly-12-hydroxy stearic acid of ≈15 nm thickness.The results challenge aspects of the data andtheoretical explanations in the current literature, andreveal a crucial role played by alkane impurities.

Rheological data suggest that, contrary to previousexpectation, network formation is possible withparticles over a broad range of sizes. Only littledependence on particle size is observed. NMR dataindicate the presence of a significant fraction ofisotropic material down to 10 K below the bulknematic transition of the liquid crystal. Calorimetricfindings suggest that small amounts of alkaneimpurities, carried originally by the dispersedparticles, are present. These molecules act as a second solvent. They play a crucial role in ‘tuning’the kinetics of phase separation to allow networkformation by (i) opening up a biphasic region and (ii) wetting the particles with a layer of isotropicmaterial.

Condensed Matter:Top Papers 2004 Showcase

Magnetic force micrographs of a multiwalled nanotube lying on a

magnetite substrate obtained in the amplitude mode with a low-moment,

permalloy-coated tip.

Probable crystal structure for KOs2O6. K, Os and O atoms occupy the

8b, 16c and 48f sites in the space group Fd �3m, respectively.

(a) A 1H NMR solid echo spectrum, taken at 30 °C (particle size: R

= 430 nm). (b) The temperature dependence of the fraction of the

isotropic phase for different particle sizes. ■ : R = 120 nm; ▼▼: R =

430 nm; �: R = 780 nm. Weight fraction of colloids: 15%.


Unexpected atom exchange insurface diffusionNovel exchange mechanisms in the surface diffusion of oxidesDuncan J Harris, Mikhail Yu Lavrentiev, John H Harding, Neil L Allan and John A PurtonJ. Phys.: Condens. Matter 16 L187-L192

Usually, one expects surface diffusion to occur simplyby adatoms moving around on top of the surfaceatoms. Surprisingly, in some oxides, the diffusingspecies move most rapidly by exchanging withsubstrate atoms. This behaviour, unexpected inoxides, showed up through new methods ofsimulation. Work by John Harding (UCL) and NeilAllen (Bristol) and colleagues shows the importanceof exchange mechanisms in surface diffusion onoxides. Simulation of diffusion in simple ceramicspresents particular problems because energy barriersare often high, so it is impossible to run moleculardynamics simulations for long enough to obtainadequate statistics. Temperature accelerateddynamics (TAD) uses simulations performed at hightemperature to calculate the evolution of systems at a lower temperature of interest.

The authors have simulated the diffusion of BaOmolecules along the (100) surface of a BaO substrate.The TAD calculations clearly show that the diffusionmechanism with the lowest barrier involves exchangewith surface ions.

They have also calculated the mean formation timeof islands on the surface, starting from a randomdistribution of molecules. As the simulation proceeds,randomly moving pairs meet and create initialclusters, which cannot move; the proto-steps soproduced attract molecules in turn, resulting in thespontaneous creation of islands. Including theexchange mechanism drastically accelerates islandformation. The evolution of the islands is shown inthe figure.

Molecular beam epitaxy is being used to createlayered structures of ferroelectric, ferromagnetic and dielectric oxides. The question of whether theexchange mechanism is still active when a moleculeof one oxide diffuses on a different oxide substrate isof fundamental importance for creating sharpinterfaces in such structures. The authors haveperformed calculations using a BaO molecule on the (100) SrO surface and SrO on the (100) BaOsurface. The large size of the barium ion, relative tothe strontium ion, prevents embedding of the Ba ioninto the SrO surface, whereas the Sr ion readilyexchanges with a surface layer barium ion. In thiscase the activation energy is about 0.6 eV. Theexistence of this low-energy mechanism suggeststhat, in some cases at least, ionic materials cannot begrown on a substrate with a similar structure withoutsignificant intermixing.

The authors have shown the importance ofexchange mechanisms in a variety of surfacediffusion processes in simple oxides. The mixingeffect inherent in such a process means that thepossibility of these mechanisms must be consideredwhen attempting to build oxide nanostructures andmultilayers.

Surface recrystallization by electronsElectron-stimulated athermal surface recrystallization of Si(100)Tetsuya Narushima, Masahiro Kitajima and Kazushi MikiJ. Phys.: Condens. Matter 16 L193-L200

Electronic excitation is a remarkably effective tool forcontrolled modification of materials. Even low-energy electron beams, typically under 5000 eV,routinely used to investigate the structural andchemical properties of surfaces, can cause structuralchanges. Tetsuya Narushima and colleagues atNational Institute for Materials Science have found a more surprising result: complete restoration of thedisorder-induced surface stress of Si(100) occurs forelectrons of low energies 3.75-40 eV. Uponirradiating the disordered Si surface, disorder-induced compressive stress completely relaxed. Thestress relaxation was found to depend only on thenumber of irradiated electrons and was independentof the total energy deposition, indicating anunderlying mechanism that is athermal. They suggestthat there is a threshold around 40-90 eV (in the caseof Si) between athermal and thermal electron-surface atom interaction processes in the surfacelayer. At energies below the threshold, the electronscan heal the damage, for example, caused by low-energy Ar+ ions.

In a recent letter, Narushima et al show usingscanning tunneling microscopy (STM) that the originof the athermal surface stress relaxation isrecrystallization of the surface atoms. In general,higher-energetic particles are expected to penetratedeeper inside materials, but electrons with extremelylow kinetic energy ��70 eV) can penetrate tounusual depths because they cannot lose energy viaany energy loss process. Higher-energy electronirradiation (> 90 eV) induces surface defects in the Sisurface by electron excitation due to the cascade ofinelastic scattering events. Narushima’s extremelylow-energy electron irradiation achieves restorationof a destroyed structure. A plausible mechanism ofthe athermal restoration observed is ionization-enhanced diffusion via the charge state transition ofdefects where no energy loss occurs. In defective Si,electron irradiation promotes athermal diffusion ofdefects. Such a charge state transition mechanismmay occur on the Si surface. Some traces of theathermal restoration on the surface structure couldbe observed, such as a slightly pairing (1�1) structureand an absence of buckled dimers and out-of-phasedimerization features derived from the athermalprocess of electron irradiation. In demonstrating thatsurfaces can have the capacity of self-healing, theseresults suggest new opportunities for atomic-scalesurface engineering.

Non-linear response to ultra-shortlaser pulsesResponse theory for time-resolved second-harmonic generation and two-photonphotoemissionC Timm and K H BennemannJ. Phys.: Condens. Matter 16 661-694

It is now possible to create ultrashort (a fewfemtoseconds) laser pulses, whose duration is similarto the relaxation times of excited electrons andcollective excitations in solids. These enable non-equilibrium physics of condensed matter systems to be studied using nonlinear techniques such as time-resolved sum frequency generation (SFG) andtwo-photon photoemission (2PPE).

A recent paper by Carsten Timm and KarlBennemann of Free University Berlin presents aunified response theory for the time-resolved second-harmonic generation (SHG) and 2PPE and derivesthe dependence of the SFG light intensity and the2PPE photoelectron yield on the time dependence ofthe exciting laser field. The theory does not rely onany assumption about the time or frequencydependence of the exciting laser pulses. The paperdiscusses metals but the response theory can beapplied to semiconductors and insulators as well.

They study a simple tight-binding model of a metalto show that the theory gives reasonable numericalresults and to illustrate effects important for theunderstanding of SFG and 2PPE. They show howrelaxation rates and detuning affect the interferencepatterns in single-colour pump-probe SHG and2PPE experiments: the lifetime in the intermediatestates and their detuning with respect to the photonenergy lead to a similar narrowing of the interferencepatterns. The effect of detuning must be taken intoaccount in order to extract meaningful lifetimes fromsuch experiments. Also, in particular in SHG themeasured relaxation rate is a weighted average overthe relaxation rates of many excited states.Furthermore, the weights in this average change withthe pump-probe delay. Thus different rates governthe decay of the interference pattern depending onthe pump-probe delay-the decay is not simplyexponential.

See also a review by Bennemann

Ultrafast dynamics in solidsK H BennemannJ. Phys.: Condens. Matter 16 R995-R1056

This review analyses the ultrafast response of metalsand semiconductors to electronic excitations.


Evolution of BaO islands on the (100) surface of BaO. (a) Starting

configuration; (b) after 380 ns simulation at 1000 K.

Detailed comparison between (a) a typical thermally annealed surface

and (b) an electron-irradiated surface after ion bombardment. Out-of-

phase dimers are seen in the broken circle.

A simplified representation of sum-frequency generation (SFG). EF is

the Fermi energy and Evac is the vacuum energy. In SFG two photons of

frequencies �1 and �2 are absorbed by electrons in states |1� and

|2� and a single photon of frequency �1 + �2 is emitted due to the

electronic transition |3�➝ |1�. The two photons may be provided

by one or two laser pulses. Note that whether the two photons are

predominantly absorbed at nearly the same time or with some delay

depends on the shape and width of the pulse(s).


Simulation of highly disorderedsolidsSimulation of mineral solid solutions at zero and high pressure using lattice statics, lattice dynamics and Monte Carlo methodsI T Todorov, N L Allan, M Yu Lavrentiev,C L Freeman, C E Mohn and J A PurtonJ. Phys.: Condens. Matter 16 S2751-S277

Grossly disordered minerals and many oxides ofmajor practical importance, such as oxide nuclearfuels, present major theoretical challenges. Thesesystems may contain large numbers of defects, ormay be non-ideal solid solutions. Neil Allen (Bristol)and colleagues discuss two practical multi-configurational ways to handle such systems withfinite impurity or defect concentrations, far from the dilute limit.

The first builds on a highly efficient method for the fully dynamic structure optimization of large unitcells which uses lattice statics and quasiharmoniclattice dynamics (QLD). The accurate calculation of the free energy via QLD is quick andcomputationally efficient and does not resort tolengthy thermodynamic integration. The full set offree energy first derivatives is calculated analyticallyand a minimization of the free energy with respect toall structural variables for large unit cells is possible.Here this technique is extended to evaluate the freeenergies of solid solutions (including ▲▲Hmix, ▲▲Smix)and phase diagrams at any pressure. This is achievedby forming a thermodynamic average of the freeenergies of a number of configurations. Strategies for generating a suitable set of configurations arediscussed. We compare results obtained by randomgeneration with those obtained using radialdistribution functions or explicit symmetryarguments to obtain approximate or exact weightingsrespectively for individual configurations.

The second technique is the well-known MonteCarlo method, extended in such a way that both theatomic configuration and the atomic coordinates ofall the atoms are changed. This approach is calledMonte Carlo exchange (MCX). While absolutevalues of the free energy cannot be obtained readilyfrom Monte Carlo simulations, the semigrandcanonical ensemble provides a convenient route toaccurate chemical potential differences accurately andhence the phase diagram.

Both methods are readily applied to high pressuresand elevated temperatures without the need for anynew parametrization. Agreement between the twotechniques is better at high pressures whereanharmonic terms are smaller. The authors comparein detail the use of each technique for properties suchas enthalpies, entropies, volume and free energies ofmixing at zero and high pressure and thus calculationof the phase diagram. They assess the vibrationalcontributions to these quantities and compare resultswith those in the dilute limit. The techniques areillustrated throughout using MnO-MgO and shouldbe readily applicable to more complicated systems.

Glass transition of polymer filmsGlass transition temperatures of polymer thinfilms monitored by Raman scatteringH Liem, J Cabanillas-Gonzalez, P Etchegoinand D D C BradleyJ. Phys.: Condens. Matter 16 721-728

Donal Bradley’s group at Imperial College Londonhave developed a new method for monitoring theglass transition in polymer thin films, which is oftechnological importance in polymer devices and aninteresting problem of fundamental physics. Belowthe glass transition the polymer behaves as a harddisordered solid with a microscopic glassy structure,while above it the polymer becomes rubber-like andcomparatively softer.

It is known that the effective glass transition in thinfilms depends strongly on the film thickness and theinterfacial interactions. The glass transitiontemperature of a thin polystyrene film on a Sisubstrate was found to be depressed by ~20°Cwhereas in a free-standing film with the samethickness it was depressed by ~70 °C, clearly showingthe role of interfacial interactions.

It has been suggested that the boundary of the film has a surface layer of finite thickness with anincreased mobility, producing a lower Tg. This high-mobility layer increases as the temperatureapproaches the bulk Tg upon heating. The origin ofthe increased mobility of the polymer chains at thesurface layer is still not clear.

Previously, Brillouin scattering has been used tomonitor Tg, but Bradley introduced the use ofconfocal Raman spectroscopy. Experiments wereperformed on free-standing polystyrene films withvarying thickness as a model study to validate thetechnique. Excellent agreement with previousdeterminations of Tg by other techniques was found.

The technique offers further possibilities, which arenot fully exploited in the case of polystyrene owing toits simple nature. As pointed out before, polarizedscattering in polystyrene gives essentially the samequalitative information for both I� and I�, but this isnot necessarily the case in more complex polymersthat may display molecular order at the surfaces. Theauthors have found evidence of this in thin films of aconjugated polymer. They also found evidence of acrossover in which two glass transitions are observedif REE < d (REE = polymer chain length and d =thickness of the film) but only a single transition isobserved if REE > d. A simple interpretation of this isthat, for REE > d, the distinction between the surfaceand the bulk of the film becomes impossible to makeand a single transition is expected accordingly.

All these results are part of a more comprehensivestudy of glass transitions in conjugated polymer thinfilms which is underway.

Making a diode from isomersThe heteroisomeric diodeAnthony N Caruso, Ravi B Billa, SnjezanaBalaz, Jennifer I Brand and P A DowbenJ. Phys.: Condens. Matter 16 L139-L146

Semiconductor diodes are based on a junctionbetween p- and n-type materials. They are usuallyeither heterojunctions (made of two differentsemiconductors) or homojunctions (a singlesemiconductor doped to form p-type and n-typeregions). The heteroisomeric diode is a new type ofdiode formed between two polytypes of the samematerial where one polytype is effectively p-type and the other n-type. This is possible if thesemiconductor can exist in different polytypes (i.e.different atomic arrangement with the same chemicalcomposition).

Peter Dowben’s group at University of NebraskaLincoln have made heteroisomeric diodes bychemical vapour deposition (CVD) from twodifferent isomers of closo-dicarbadodecaborane(C2B10H12) - which tend to form p- and n-typesemiconducting boron carbide respectively. Thesecarborane heteroisomer diodes have manyadvantages over the previously fabricatedheterojunction and homojunction diodes. Theisotope 10B has a large neutron capture cross-section,so diodes with vastly improved sensitivity can bemade by using 100% 10B (its natural abundance isabout 20%). This can be achieved efficiently via 10Benrichment of the carborane source molecules usedfor CVD. Furthermore, since the heteroisomericdiodes are formed without transition metal dopants,activatible transition metals can be eliminated,leading to increased reliability and fewer false positivesignals.

Neutron detection efficiencies of these new diodeshave yet to be measured, and they have not, as yet,exhibited the insensitivity to high temperature foundwith the silicon carbide/boron carbideheterojunction and the boron carbide homojunctiondiodes. Nonetheless, the heteroisomeric diodespackaged in organic polymers (that also act as aneutron moderator) are more efficient and subject tofewer false positives than some recent schemes fordetecting fissile materials. Even without isotopicenrichment, since the depletion regions and bothsides of the p-n junction contain a boron-richsemiconductor, the heteroisomeric diodes represent a potentially far more efficient (and hence sensitive)solid state neutron detector than any conversion layeror heterojunction layer device.


Calculated values of ▲▲Hmix (kJ mol_1) for a 50/50 MnO-MgO

composition at 1000 K as a function of supercell size and number

of configurations. The number attached to each curve denotes the total

number of atoms in each supercell.

Parallel Raman intensity (I�) as a function of T of polystyrene for

different film thicknesses going from 180 to 50 nm. A distinct crossover

is observed below a specific thickness, which depends on the molecular

weight.

The HOMO-LUMO gap of orthocarborane (C2B10H12) (top) and

metacarborane (C2B10H12) (bottom). Shown are the combined pho-

toemission (left) and inverse photoemission (right) spectra of the molec-

ular thin films on gold and copper for metacarborane and orthocarborane

respectively. The highest occupied molecular orbital (HOMO) and

the lowest unoccupied molecular orbital (LUMO) are identified from

detailed coverage-dependent studies and compared with the theoretical

expectations for the isolated molecule. Blue atoms are carbon and red are

boron. The Fermi level is closer to the binding energy of the HOMO for

orthocarborane and closer to the LUMO for metacarborane.


Anisotropy in magnetic semiconductorsMagnetic anisotropy and domain structure in carrier-controlled ferromagnetic semiconductorsTomasz DietlJ. Phys.: Condens. Matter 16 S5471-S5479

There is an intense search for room-temperaturespintronic semiconductor materials. Much effort hasbeen devoted to understanding the nature of carrier-controlled ferromagnetism in tetrahedrallycoordinated diluted magnetic semiconductors(DMS), such as (Ga, Mn)As. A theory based onZener’s model of ferromagnetism, the Ginzburg-Landau approach to the phase transitions, and theKohn-Luttinger kp theory of semiconductorsaccounts for the magnetic properties of Mn-dopedGaAs, InAs, GaSb, and InSb as well as in p-CdTe,p-ZnTe, and Ge, assuming that the long-rangeferromagnetic interactions between the localizedspins are mediated by the holes in the weaklyperturbed valence band. The understanding of(III, Mn)As alloys has provided a basis for thedevelopment of novel methods enablingmagnetization manipulation and switching.

Tomasz Dietl of the Polish Academy of Sciencesreviews the experimental situation and theoreticalmodelling of the micromagnetic properties of (Ga,Mn)As and related compounds. Interestingly, despitemuch lower spin and carrier concentrationscompared with ferromagnetic metals, these materialsexhibit excellent micromagnetic characteristics,including well defined magnetic anisotropy and largeand ferromagnetic domains separated by usuallystraight-line domain walls.

The Zener p-d model explains quantitatively theeffect of strain on the easy axis direction as well aspredicting correctly the presence of the reorientationtransition, observed as a function of holeconcentration and temperature. Possible suggestionsto explain the existence of in-plane uniaxialmagnetocrystalline anisotropy are put forward.Finally, magnetic stiffness computed within the samemodel of ferromagnetism is presented, and is shownto explain the domain width in perpendicular films.It is emphasized that a rather large magnitude of thestiffness accounts for both the excellentmicromagnetic properties of ferromagnetic (III, Mn)V films and the quantitative applicability of the mean-field approximation to these materials.

Energy landscape of glassypolystyreneA study of the deep structure of the energy landscape of glassy polystyrene: the exponentialdistribution of the energy barriers revealed byhigh-field electron spin resonance spectroscopyV Bercu M Martinelli, C AMassa, L A Pardi and D LeporiniJ. Phys.: Condens. Matter 16 L479-L488

Solid glassy dynamics is currently an active field ofstudy. At temperatures well below the glass transitiontemperature Tg (so the ageing effect can be ignored)but high enough that the tunnelling effects governingthe low-temperature anomalies of glasses can beignored, the dynamics is thermally activated in thesubstructures of the minima of the energy landscapeaccounting for various subtle degrees of freedom. Ina glass the temperature dependence of the energybarrier distribution g(E) is only weakly temperaturedependent.

The shape of the energy barrier distribution g(E)

in glasses has been extensively investigated byexperiments, theories and simulations. It is usuallyfound to be either Gaussian or exponential. Aconvolution of these two distributions as well as the truncated Levy flight, i.e. a power law withexponential cut-off, resembling a stretchedexponential have also been considered.

D Leporini and co-workers from Pisa have studiedthe reorientation of one small paramagneticmolecule (spin probe) in glassy polystyrene by high-field electron spin resonance spectroscopy at twodifferent Larmor frequencies (190 and 285 GHz).They found unambiguous evidence for anexponential distribution of the energy barriers for therotational motion of the spin probe at both 240 and270 K. The same shape for the distribution of theenergy barriers of polystyrene was evidenced by themaster curves provided by previous mechanical andlight scattering studies. The breadth of the energybarrier distribution of the spin probe is in the range of the estimates of the breadth of the polystyreneenergy barrier distribution. The evidence that thedeep structure of the energy landscape ofpolystyrene exhibits the exponential shape of theenergy barrier distribution agrees with the resultsfrom extreme-value statistics previously found byBouchaud and co-workers.

Charged critical fluctuations in anoxide superconductorEvidence for charged critical fluctuations in underdoped YBa2Cu3O7-�T Schneider, RKhasanov, K Conder,E Pomjakushina, R Bruetsch and H KellerJ. Phys.: Condens. Matter 16 L437-L442

Close to the critical temperature Tc of the normal-superconductor transition, order parameterfluctuations dominate the critical properties. Forstrongly type I materials the coupling of the orderparameter to transverse gauge field fluctuations isexpected to render the transition first order, whereasstrongly type II materials should exhibit a continuousphase transition, and sufficiently close to Tc thecharge of the order parameter is relevant. However,in cuprate superconductors within the fluctuation-dominated regime, the region close to Tc, where thesystem crosses over to the regime of chargedfluctuations, turns out to be too narrow to access. Forinstance, optimally doped YBa2Cu3O7-�, whilepossessing an extended regime of criticalfluctuations, is too strongly type II to observe chargedcritical fluctuations. However, underdoped cupratescould open a window onto this new regime because �is expected to become rather small. Noting that Tcdecreases on approaching the underdoped limit,sufficiently homogeneous and underdoped cupratesappear to be potential candidates to observe chargedcritical behaviour.

T Schneider and co-workers in Zürich report andanalyse in-plane penetration depth measurements ofunderdoped YBa2Cu3O7-� to explore the evidencefor this new critical behaviour. These measurementsprovide consistent evidence for the charged criticalbehaviour of the superconductor-normal statetransition in type II superconductors (� > 1/√2).Since the effective dimensionless charge ̃ e = 1/�scales as Tc

-1/2 this new critical behaviour should beobservable in suitably underdoped cuprates. In thisregime the crossover upon approaching Tc is thus tothe charged critical regime, while near optimumdoping it is to the critical regime of a weakly chargedsuperfluid where the fluctuations of the orderparameter are essentially those of an unchargedsuperfluid. Furthermore, there is the inhomogeneityinduced finite size effect which renders theasymptotic critical regime and with that the chargedregime of nearly optimally doped samples difficult toattain


Curie temperatures of (Ga, Mn)N, (Ga, Mn)P, (Ga, Mn)As and (Ga,

Mn)Sb calculated from first principles in the mean field approximation.

[from Exchange interactions in diluted magnetic semiconductors, K Sato,

P H Dederichs, H Katayama-Yoshida and J Kudrnovsk_, J. Phys.:

Condens. Matter 16 S5491-S5497]

The ESR line shapes at 190 GHz (left) and 285 GHz (right) of

TEMPO in PS at 50 K. The magnetic parameters are: gx = 2.009

94 ± 3 � 10_5, gy = 2.006 28 ± 3 � 10_5, gz = 2.002 12±3

�10_5, Ax(mT) = 0.62±0.02, Ay (mT) = 0.70±0.02, Az (mT)

= 3.40±0.02. The superimposed dashed curves are best fits according

to the SCT model, with SCT = 25 ns (190 GHz) and SCT= 19 ns

(285 GHz). The jump angle = 60°. The theoretical lineshapes were

convoluted by a Gaussian with width w = 0.15 mT to account for the

inhomogeneous broadening.

(d ln �ab/dT )_1 with �ab in µm versus T for YBa2Cu3O6.59. The

straight line with slope 1/� � 3/2 corresponds to charged criticality

with Tc = 56.1 K, while the dashed line indicates the intermediate 3D

XY critical behaviour with slope 2/� � 3.


Unusual magnetic structures in Gd2Ti2O7Phase transitions, partial disorder and multi-k structures in Gd2Ti2O7J R Stewart, G Ehlers, A S Wills, S T Bramwelland J S GardnerJ. Phys.: Condens. Matter 16 L321-L326

Magnetic oxides show an enormous variety of spinstructures, magnetic phase transitions, andphenomena such as frustration. They pose challengesto theory and experiment, and force new thinkingabout cooperative phenomena. Ross Stewart of ILL,Grenoble, with co-authors from Oak Ridge, UCLand NIST, have studied the geometrically frustratedantiferromagnet Gd2Ti2O7, which exhibitsparticularly interesting and complex magneticbehaviour. Magnetic ordering commences at TN =1.1 K and there is a further magnetic phase transitionat T�= 0.7 K.

The technique of neutron diffraction enables thedirect determination of complex magnetic structures,but it can be ambiguous when applied to systems ofhigh symmetry, where the problem of ‘multi-k’structures arises. Alternative structures, described byone or several symmetry-related propagation vectorsk, give identical diffraction patterns as a result oforientational averaging in powders or ‘k-domain’formation in single crystals. To distinguish suchstructures is difficult.

The 1-k and 4-k structures of Gd2Ti2O7 are shownin the figure. In the 4-k structure, all spins (in bothsets) are perpendicular to the local trigonal ‹111› axes(that connect the vertices of the tetrahedra to theircentres), while in the 1-k structure the spins areperpendicular to a single global [111] crystallographicdirection. However, the disordered spin components—which are associated with only theweakly ordered set—have a completely differentspatial distribution in the 1-k and 4-k structures. In the1-k structure, nearest neighbour disordered spincomponents are separated by 7.2Å, while in the 4-kstructure they are separated by 3.6 Å. The magneticdiffuse neutron scattering from the two structuresmust therefore be substantially different.

Between T� and TN the structure is partly ordered,as previously reported. Below T� the remaining spinsorder, but only weakly. The magnetic structure in thistemperature range is shown to be a 4-k structure,closely related to the 1-k structure previouslysuggested. The 4-k and 1-k variants of the structureare distinguished by analysis of the diffuse scattering,which represents a new method of solving the ‘multi-k’ problem of magnetic structuredetermination.

Ferroelectric nano-capacitors Intrinsic dielectric response in ferroelectricnano-capacitorsM M Saad, P Baxter, R M Bowman,J M Gregg, F D Morrison and J F ScottJ. Phys.: Condens. Matter 16 L451-L456

Ferroelectric perovskite oxides are key materials inrandom access memory (RAM) technology, but theorigins of changes in functional properties betweenthin film and bulk remain unresolved. Ferroelectricthin films show a significant broadening of the Curieanomaly (peak in dielectric constant with respect totemperature) as films are made thinner. Thebroadening of the dielectric peak is associated withan apparent change in the nature of the paraelectric-ferroelectric phase transition from first to secondorder.

Jim Scott (Cambridge) with co-workers in Belfasthave performed measurements on ‘free-standing’single-crystal barium titanate capacitors withthickness down to 75 nm, which show a dielectricresponse typical of large single crystals rather thanconventional thin films, with no broadening ortemperature shift of the dielectric peak or losstangent. They observed peak dielectric constants of~25 000 and Curie-Weiss analysis demonstrated firstorder transformation behaviour. This is in dramaticcontrast to results on conventionally deposited thinfilm capacitor heterostructures, which show largedielectric peak broadening and temperature shifts,as well as an apparent change in the nature of theparaelectric-ferroelectric transition from first tosecond order. Their data are compatible with a recentmodel by Bratkovsky and Levanyuk, which attributesdielectric peak broadening to gradient terms that willexist in any thin film capacitor heterostructure. Theobserved recovery of first order transformationbehaviour is consistent with the absence of significantsubstrate clamping in our experiment, as modelled byPertsev et al, and illustrates that the second orderbehaviour seen in conventionally deposited thin filmscannot be attributed to the effects of reduceddimensionality in the system, nor to the influence ofan intrinsic universal interfacial capacitanceassociated with the electrode-ferroelectric interface.Rather, the influence of homogeneous strain throughstrain coupling to a substrate, or gradient termsassociated with chemical, defect or strain gradients,are the primary suspects.

Heat dissipation in nanostructuresOpen-boundary Ehrenfest molecular dynamics:towards a model of current induced heating innanowiresAndrew P Horsfield, D R Bowler and A J FisherJ. Phys.: Condens. Matter 16 No 7 L65-L72

Much of nanotechnology is driven by demands forsmaller, faster and more robust devices. Yet reducingsizes brings problems, one being heat dissipation innanoelectronic devices. To calculate heat dissipationin a nanoscale device, where most standardassumptions fail, is the challenge addressed byAndrew Horsfield, David Bowler and Andrew Fisherat University College, London. They developed atime-dependent method based on the single-particleelectron density matrix that allows the electronic andionic degrees of freedom to be modelled within theEhrenfest approximation in the presence of openboundaries. It describes a practical implementationusing tight binding, and uses it to investigate steady-state conduction through a single-atom device and toperform molecular dynamics. It is found that in theEhrenfest approximation an electric current allowsboth ionic heating and cooling to take place,depending on the bias.

To monitor ionic heating the authors follow theevolution of the kinetic energy with time (see figure).For a small bias (0.1 V) the ionic kinetic energy decayswith time (cooling), while for a large bias (1.0 V) thekinetic energy increases (heating). The energy ( �)associated with the vibrational frequency of the ion is0.055 eV, and equals the Born-Oppenheimer surfaceseparation for allowed transitions. Hence the changein bias increases approximately tenfold the number ofpossible heating transitions, producing the observedchanged behaviour. This method is thus able tocalculate some non-adiabatic effects of current flow.However, the amount of heating that is observed for abias of 1 V is much less than would be expected fromquantum perturbation theory. This is the subject ofongoing work by the authors.

See also a recent topical review by David Bowler.

Atomic-scale nanowires: physical andelectronic structureD R BowlerJ. Phys.: Condens. Matter 16 R721-R754

The technology to build and study nanowires withsizes ranging from individual atoms to tens ofnanometres has been developing rapidly over the lastfew years. This review surveys the motivation behindthese developments, and summarizes the basicsbehind quantized conduction. It describes several ofthe different experimental techniques and materialssystems used in the creation of nanowires, and therange of theoretical methods developed both forexamining open systems (especially their conductionproperties) and for modelling large systems. Itpresents various noteworthy example results from thefield, and concludes with a look at future directions.


The frequency dependence of the capacitance measured at three

different temperatures for thin BaTiO3 lamellae. All lamellae investi-

gated showed the distinct lack of spectral dependence illustrated here.

Strong temperature dependence is, however, evident.

Comparison of (left) the 1-k structure and (right) its 4-k variant.

In each structure the four Gd3+ ions coloured orange are shown as

carrying no thermally averaged moment. The phase transition at T�

= 0.7 K involves weak ordering of these four spins and a small canting

of the remaining spins away from the positions shown. At T << 0.7

K, only these ions carry a disordered spin component.

The variation of the ionic kinetic energy of the mobile atom with time

for a bias of (a) 0.1V and (b) 1.0 V. The device contains 3 atoms, and

the leads 16 atoms. The initial temperature is about 600 K.


Multiple spin-relaxation in CMR materialsThe origin of multiple spin-relaxation channelsbelow the metal-insulator transition in ferromag-netic colossal magnetoresistance (CMR)manganitesR H Heffner, D E MacLaughlin,G J Nieuwenhuys and J E SonierJ. Phys.: Condens. Matter 16 S4541-S4562

Perovskite-structured Mn-based materials show arange of interesting properties that can be controlledby doping with divalent alkaline earth elements. Ofparticular interest is ‘colossal magnetoresistance’(CMR): a reduction (�10) in resistance occurs nearthe metal-insulator transition temperature TMI whena magnetic field of a few tesla is applied. CMRmaterials are under intense investigation for theirpossible use in spintronics devices, for which spinrelaxation is an imprtant issue.

Bob Heffner (Los Alamos) and co-workers haveperformed a series of muon spin relaxation (µSR)experiments on perovskite manganites in the La1-xCaxMnO3 series. Their main focus was theobservation of two distinct Mn relaxation channels in ferromagnetic CMR materials. Conventionalhomogeneous ferromagnets exhibit only a singlerelaxation channel above and below the Curietemperature TC. As La1-xCaxMnO3 ferromagnetsnear optimal doping had been believed to reach ahomogeneous state just a few degrees below TC dueto the long-range transport of the doped holes thatgive rise to a metallic state, the initial observation ofmultiple relaxation channels significantly below TCwas puzzling.

The measurements on (La, Ca, Pr)MnO3compounds showed that ferromagnetic CMRcompounds develop two Mn-ion spin-latticerelaxation channels as they are cooled below theirinsulator-metal transition temperature TMI. Thisresult is in contrast to conventional ferromagnets andis attributed to the presence of both insulating andconducting ferromagnetic regions below TMI whichcoexist on a microscopic scale. The coexistence ofdifferent phases is found in both polycrystalline andsingle-crystalline materials, though the single crystalexhibits a narrower temperature region of phasecoexistence below its ferromagnetic criticaltemperature TC. The authors discuss possibledifferences between crystalline and granularmaterials which could give rise to these findings.These results could have important implications for the use of CMR materials in spintronics devices,which rely on conducting surfaces in thin filmmultilayers below TC, where even a thin layer ofinsulating surface material can degrade performance.

Fatigue crack closuremicromechanismsIn situ high resolution synchrotron x-ray tomog-raphy of fatigue crack closure micromechanismsK H Khor, J-Y Buffière, W Ludwig, H Toda,H S Ubhi, P J Gregson and I SinclairJ. Phys.: Condens. Matter 16 S3511-S3515

Fatigue crack closure can have major effects on crackgrowth rates, but the methods of measuring it arecontroversial. To date, computed finite-elementmodels, analytical models and widely establishedcompliance-based experimental methods haveoffered limited micromechanical insight and/ordirect information on the active crack tip regionwithin bulk material. To understand the absolutecontributions of crack closure mechanisms, such asplasticity-induced and roughness-induced closure,to fatigue properties, an internal, three-dimensionalinsight into crack behaviour during loading andunloading is clearly of value.

K H Khor (Southampton) and co-workers havecarried out synchrotron radiation x-raymicrotomography at a high resolution of 0.7 µm toprovide unique three-dimensional in situ observationof steady state plane strain fatigue crack growth in a2024-type Al alloy (Al-Cu-Mg-Mn). Using such highresolution imaging (additionally exploiting the phasecontrast effect in interface imaging), the details offatigue cracks are readily observed, along with theoccurrence of closure. They used a novelmicrostructural crack displacement gauging methodto quantify the mixed mode character of crackopening displacement and the closure effect. Theyobserved details of a crack, such as its surfacetopology, bifurcation and tip geometry, by volumerendering, and obtained evidence of the occurrenceand mechanical influence of crack closure. Theyused a liquid gallium grain boundary wettingtechnique in conjunction with the microtomographyto visualize the correlation between the three-dimensional structure of the grains and fatigue crackbehaviour. Subsequently, electron backscatteringdiffraction assessment of the grain orientation on thesamples provided a uniquely complete 3Ddescription of crack-microstructure interactions.

Modelling Charged colloidal suspensionsFluid particle dynamics simulation of chargedcolloidal suspensionsHiroya Kodama, Kimiya Takeshita,Takeaki Araki and Hajime TanakaJ. Phys.: Condens. Matter 16 L115-L123

The most difficult problem in studying the dynamicsof charged colloidal suspensions arises from thecomplex dynamic coupling between the motions ofthe three key elements, namely colloidal particles,ions, and liquid molecules. These elements interactstrongly with each other via both electrostatic andhydrodynamic interactions, which are both of long-range nature, so this is a very complex dynamicmany-body problem. In most previous studies, themotion of one of these elements has been ignored.This study by Hajime Tanaka’s group is the first totake all three components into account.

In principle, molecular dynamics (MD) simulationis a powerful means of studying the dynamicbehaviour of charged colloidal suspensions. Itincorporates hydrodynamic and electrostaticinteractions by explicitly solving for the motions of colloids, ions, and liquid molecules. In reality,however, it is too costly in computer time for dealingwith the hydrodynamic motion of liquid moleculesand the resulting motion of colloids and ions, soliquid molecules have previously been neglected. Thenumber of colloidal particles and the timescale haveto be limited somehow: coarse graining of theproblem is essential.

A new simulation method (the fluid particledynamics (FPD) method) was recently proposed forcharge-free colloidal suspensions by Tanaka andAraki. Here liquid molecules are treated as acontinuum fluid and a solid colloidal particle istreated as an undeformable fluid of high viscosity.This allows the whole colloidal system to be treated as a continuous liquid with inhomogeneity of theviscosity, which exactly reflects the spatial distributionof colloidal particles.

In the present letter, Tanaka and co-workerspresent a new simulation method to deal with thedynamics of charged colloidal suspensions whileincluding full hydrodynamic and electrostaticinteractions among charged colloidal particles, ionclouds, and liquid. The validity of this method hasbeen demonstrated for the problem ofelectrophoretic deposition kinetics. The electrostaticand the hydrodynamic interactions as well as theelectro-osmotic effect are naturally introduced in thesimulation. Although they have shown a two-particlesimulation of a small system size in two dimensions,it is quite straightforward for their method to treat amany-particle system and/or a three-dimensionalsystem. The new method should contribute to adeeper understanding of kinetic aspects of complexmany-body problems in charged colloidalsuspensions.


Temperature dependence of the resistivity for class I La1-xCaxMnO3 for

x = 0.0, 0.06, 0.18 and 0.33.

Comparison of the EBSD grains map and tomography data: (a) 3D

reconstruction of grain boundaries (yellow) and the crack (green),

(b) grain misorientation data from the EBSD data in (c), along the line

y-y, and (c) an EBSD orientation map of the sample at the front plane

of the image (a).

Snapshots of the process of the electrophoretic deposition of two charged

particles placed between the electrodes. The density plot represents the

counterion concentration. The arrows indicate the velocity field. The elec-

trodes are separated by 310 nm (d = 310 nm). The applied voltage is

V = 100 mV. The average concentration of the added salt is 10-5 M.


Wall tensions of colloid-polymermixturesWall tensions of model colloid-polymer mixturesPaul P F Wessels, Matthias Schmidt andHartmut LöwenJ. Phys.: Condens. Matter 16 L1-L8

Mixtures of sterically-stabilized colloidal particlesand non-adsorbing globular polymers suspended inan organic solvent are valuable soft matter systems tostudy demixing phase transitions and wettingphenomena at walls. Controlling the wettingbehaviour is mandatory for tailoring wall coatingswith intriguing applications like self-cleaningsurfaces.

The Asakura-Oosawa-Vrij (AO) model of hardsphere colloids and ideal polymer spheres is a widely-used reference system. Hartmut Löwen and co-workers at Düsseldorf and Utrecht present ananalytical expression for the wall-fluid tension ofthe AO model at a hard wall using ideas from scaledparticle theory. The key idea is to consider a ternarysystem of colloids, polymers and a dummycomponent of large size and vanishing concentrationthat is equivalent to a planar wall.

They have derived analytical expressions for thewall tension of a colloid-polymer mixture using ascaled particle approach. They considered both theAO model near a hard and a polymer-coated wall(the latter being penetrable for the polymers) and thecase of interacting polymers at a hard wall. Theirresults showed good overall agreement with explicitDFT calculations. Applied to the fluid demixingbinodal, their data confirm the wetting transition atthe hard wall which has been observedexperimentally. The previously found layeringtransition manifests itself as a tiny kink in theinterfacial tension. They predict complete dryingnear a polymer-coated wall, as has been reported inexperiments. Future work could be devoted toexploring the interfacial tensions by direct computersimulation via integrating the anisotropy of thepressure tensor, by thermodynamic integration or,possibly, by using grand-canonical configurational-bias Monte Carlo to test the predictions. It would alsobe interesting to study patterned surfaces andstructured walls, which for example drasticallyinfluence the flow through microfluidic devices.

Training effect in CMR manganitesRelationship between the onset offerromagnetism and the training effect in CMRperovskite manganitesD Zhu, V Hardy, A Maignan and B RaveauJ. Phys.: Condens. Matter 16 L101-L107

Mixed-valent perovskite manganites display manyextraordinary properties including colossalmagnetoresistance and charge ordering.Pr0.5Ca0.5MnO3 exhibits orbital and charge ordering(OO/CO), which corresponds to the 1:1 ratio ofMn3+:Mn4+ stripes and is a charge-exchange-typeantiferromagnetic insulator (AFMI). This state is verystable, so a high critical field up to 25 T is needed tomelt it into a ferromagnetic (FM) state. If the Mn3+

content is increased, the OO/CO state is destabilizedand the critical field decreases. A decrease of criticalfield can also be realized by substituting foreigncations at either the A- or B-site, which can lead to thedevelopment of FM in the AFM matrix, inducingphase separation.

A Maignan and colleagues at ENSI, Caen havepreviously performed magnetic-field-drivenmagnetization measurements on these materials andfound that the AFM-FM transition shows abrupt Msteps at low temperature. This is not compatible withconventional metamagnetism, while a martensitic-like mechanism was plausible. This interpretationwas based on observations such as the training effect,influence of the measuring procedure, recovery ofthe virgin properties after annealing, role of themicrostructure, etc. The existence of phaseseparation can lead to very peculiar behaviour suchas the effect of thermal cycling, which can induce aspectacular increase of resistivity and a change in theM(H) shape. Below the critical field H1 for theappearance of the first M(H) step, M shifts down andH1 shifts up as the cycle number increases. Thisindicates that the FM regions tend to disappear to thebenefit of the OO/CO AFM ones, as if the thermalcycling was a ‘training’ effect for the stabilization ofthe AFM phase.

In a recent letter Maignan et al report on a preciseinvestigation of this so-called training effect carriedout on A-site-substituted perovskite manganites,which allows a fine tuning of the magnetic groundstate. Combining ac susceptibility and magnetizationmeasurements during thermal cyclings, they showthat the training effect is closely related to thetransition from the orbital/charge ordered state tothe ferromagnetic metallic state at low temperature,and consequently depends on TC. They propose thatthis phenomenon originates from the structural phaseseparation that appears below TC, in agreement withthe martensitic-like scenario.

Time and length scaling in spin glassdynamicsTime and length scales in spin glassesL Berthier and A P YoungJ.Phys: Condens. Matter 16 S729-S734

The rich dynamical behaviour of spin glasses hasbeen extensively studied in experiments, simulations,and theoretically. In this paper, L Berthier and A PYoung discuss the slow, nonequilibrium, dynamics of spin glasses in their glassy phase. Experimentalstudies of the spin glass phase always probe thenonequilibrium dynamics, because the equilibrationtime of macroscopic samples is infinite. Simulationscan probe equilibrium behaviour for very moderatesizes only, so the thermodynamic nature of the spinglass phase is still a matter of debate. Although manytheories account for the simplest experimental results,such as the ageing phenomenon, early experimentsrevealed several other spectacular phenomena(rejuvenation, memory, etc) that are harder toexplain.

Most previous studies have used the Edwards-Anderson model of an Ising spin glass, but very fewissues have been settled, notably the existence, for d ≥3, of a second-order phase transition to a spin glassphase. The nonequilibrium dynamics of the Isingspin glass has also been quite extensively studied but,for d = 3, some key experimental observations are notreproduced, though simulations in d = 4 have beenmore successful. This may not be too surprising, sincereal spin glasses are made not of Ising spins but vectorspins. When the interaction between spins isisotropic, the system is therefore best described by theHeisenberg spin glass Hamiltonian, treating the spinsas three-component vectors of unit length.

Very recent experiments have shown substantialdifferences between Ising (i.e. very anisotropic) andHeisenberg samples, the nonequilibrium effectsbeing much clearer in Heisenberg samples. Hence, itis hoped that dynamic studies of the Heisenberg spinglass in d = 3 will reproduce the key experimentaleffects, so that deeper theoretical knowledge of thenature of the nonequilibrium dynamics of spinglasses can be gained. This paper presents somepreliminary results from the first large-scalenumerical simulation of the nonequilibriumdynamics of the three-dimensional Heisenberg spinglass.

The results show that spin variables qualitativelyfollow the same type of ageing behaviour as in theIsing case, which is due to the slow growth with timeof a dynamic coherence length. The observation thatvery large length scales can be reached in thenumerical time window gives us the hope, alsoconfirmed by preliminary work, that the model willallow us to reproduce most of the experimentaleffects, with the advantage that simulations havedirect access to the distributions of length scalesinvolved in phenomenological theories, providingfurther understanding of spin glass dynamics.


The orientation variable cos �i (t) = Sa

i(t) · Sb

i (t) is encoded in a

greyscale in a 60�60�60 simulation box at three different times tw= 2, 27, and 57 797 (from left to right) and temperature T = 0.04.

The growth of a local random ordering is evident.

(a) Isothermal M(H) curves of Pr0.6Ca0.34Sr0.06MnO3 registered at

T = 2.5 K after different thermal cyclings (300 K-2.5 K and 300

K-2.5 K-300 K-2.5 K). (b) The enlargement of the field increasing

branches of the M(H) curves after different thermal cyclings. The

arrows indicate the magnetic field increasing and decreasing branches.

The hard wall-fluid interface tension ̃�wf = ��c2 �wf of the AO model

for size ratio q = �p/�c = 0.6 as a function of the colloid packing frac-

tion �c and for increasing polymer reservoir packing fractions

�pr = 0, 0.2, 0.4, 0.5, 0.6, 0.7, 0.8 (from bottom to top). The results

from SPT (full curves) are compared with those of numerical DFT cal-

culations (dashed curves). The gap in the top four curves corresponds to

fluid phase coexistence. Also shown is the DFT result for the polymer-

coated wall for �pr = 0.4 (dotted curve); the corresponding SPT result

is equal to �hs (lowest full curve). The inset shows the fluid part of the

phase diagram for q = 0.6 as a function of �c and �pr ; indicated are

the binodal (thick curve), spinodal (dotted curve), critical point and (hor-

izontal) tielines connecting coexisting states.


Novel quantum Monte Carlotechnique with linear scalingLinear-scaling quantum Monte Carlo techniquewith non-orthogonal localized orbitalsD Alfé and M J GillanJ. Phys.: Condens. Matter 16 L305-L311

Quantum Monte Carlo (QMC) methods, arguablythe closest approach to first principles in electronicstructure calculations, have now tackled somesignificant problems in condensed matter. Recentapplications include the reconstruction ofsemiconductor surfaces, the energetics of pointdefects in insulators, optical excitations innanostructures, and the energetics of organicmolecules. Although its demands on computer power are much greater than those of widely usedtechniques such as density functional theory (DFT),its accuracy is also much greater for most systems.With QMC now being applied to large complexsystems containing hundreds of atoms, a major issueis the scaling of the required computer effort withsystem size. In other electronic-structure techniques,including DFT, the locality of quantum coherencesuggests that it should generally be possible to achievelinear-scaling, or O(N) operation, in which thecomputer effort is proportional to the number ofatoms N. Very recently, a procedure has beensuggested by A J Williamson et al for achieving at leastpartial linear scaling for QMC, based on the idea of‘maximally localized Wannier orbitals’ (MLWO).

Here, Dario Alfé and Mike Gillan of UniversityCollege London propose and test a simpleralternative method, which appears to have importantadvantages. They have reformulated the QMCtechnique so that a large part of the calculation scaleslinearly with the number of atoms. Thereformulation is related to MLWO, but has theadvantage of greater simplicity. Their techniquedraws on methods recently developed for linear-scaling density functional theory. They report tests ofthe new technique on the insulator MgO, and showthat its linear-scaling performance is somewhat betterthan that achieved by the MLWO approach.Implications for the application of QMC to largecomplex systems are pointed out.

In addition to being simpler and more robust thanthe earlier technique, the new technique appears alsoto be more efficient. It makes it possible to treat largesystems that would be out of reach of conventionalQMC methods. Research areas where the techniqueis immediately applicable include defects andsurfaces of oxide materials, and molecular processeson these surfaces.

Watching diamond formIn situ x-ray diffraction of graphite-diamondtransformation using various catalysts underhigh pressures and high temperaturesWataru Utsumi, Taku Okada,Takashi Taniguchi, Ken-ichi Funakoshi,Takumi Kikegawa, Nozomu Hamaya and Osamu ShimomuraJ. Phys.: Condens. Matter 16 S1017-S1026

Synthesis of diamonds using industrial high-pressuretechnology is now routine, but many questions, suchas the role of the catalyst, kinetics of the reaction andthe possibility of a metastable phase, remain. Therole of inorganic compounds such as carbonates andhydroxides as catalysts for forming diamonds is also of interest. Real-time in situ experiments in which thediamond formation process is directly observedunder high pressures and temperatures areindispensable. These experiments are challengingbecause diamonds are formed in a thick and heavyhigh-pressure vessel, but are possible using verypowerful x-rays from a synchrotron radiation source.

Watari Utsumi at JAIST and co-workers haveperformed in situ x-ray diffraction studies ofgraphite-diamond transitions with various solventcatalysts under high pressures and high temperaturesat the Photon Factory and SPring-8, which combinedsynchrotron radiation and a large-volume multi-anvilhigh-pressure apparatus. They overcame manytechnical difficulties and made successful real-timeobservations of the graphite-to-diamond conversionprocess using various solvent catalysts such asconventional transition metals, carbonate materialsand aqueous fluids.

Figure 1 shows a schematic illustration of the in situ x-ray diffraction system combined with a DIAtype multi-anvil high-pressure apparatus. Theincident x-ray beam collimated by the front slitspasses through the anvil gap and then irradiates thesample in the high-pressure cell. A typical slit size is0.05 mm in width � 0.30 mm in height. A puregermanium solid-state detector mounted on agoniometer collected the diffracted x-ray collimatedby the receiving slits. The diffraction angle 2� is fixedat an appropriate value, which is selected for the bestdiffraction profiles in the region under investigation.

In Utsumi et al’s paper, the experimental procedureis described and the technical details are explained.The diffraction data with various catalysts are shownand the problems and limitation of this method arediscussed.

Phase transitions in Quasi-2D structuresObservation of a semimetal-semiconductorphase transition in the intermetallic ZrTe5D N McIlroy, S Moore, Daqing Zhang,J Wharton, B Kempton, R Littleton,M Wilson, T M Tritt and C G OlsonJ. Phys.: Condens. Matter 16 L359-L365

(Quasi)-low-dimensional materials show interestingproperties such as superconductivity, colossalmagnetoresistance and charge density wave (CDW)formation, typically accompanied by a metal-nonmetal or semimetal-nonmetal transitionassociated with either the formation of a gap orchanges in the width of the gap. Low-dimensionalsystems are inherently unstable due to correlationand exchange effects, so they exhibit structuralinstabilities such as Jahn-Teller or CDW distortions.

One class of such materials is the pentatellurides(ZrTe5 and HfTe5), which exhibit anomaloustransport properties attributed to their quasi-two-dimensional structure. Their structure consists ofzigzag chains of Te atoms along the a-axis that linkprismatic chains of ZrTe6 along the c-axis, whichtogether form 2D planes weakly bonded via van derWaals forces along the b-axis. This structure belongsto the Cmcm (D2h

17) space group.D N McIlroy of University of Idaho and co-

workers used the technique of high-resolution angle-resolved photoelectron spectroscopy to study thetemperature dependent electronic structure ofZrTe5. In the temperature range 20-170K theyobserved a band gap, indicative of a semiconductingphase. In addition, the chemical potential wasobserved to shift downwards with temperature,consistent with an intrinsic semiconductor. SinceZrTe5 is metallic for T ≤ 4.5 K, as determined fromthe Shubnikov-deHaas effect, they concluded that ametal-semiconductor phase transition occurs for 4.2K < T < 20 K. While they cannot exclude chargedensity wave formation for T < 20 K, the lack ofexperimental evidence argues against thisexplanation for the metal-semiconductor transition.However, their observations suggest that it is areduction in interplane interactions between the a-c

planes of ZrTe5 that is responsible for the transition.A thorough study of the temperature dependent

phonon density of states is needed to shed light on theorigin of the phase transition.


A schematic illustration of the in situ x-ray diffraction system combined

with a DIA type multi-anvil high-pressure apparatus.

Intensity maps of the density of states of ZrTe5 acquired along the

�-X high symmetry direction for (a) T = 170 K and (b) T = 75 K,

and (c) an intensity difference map between the maps in (a) and (b)

Convergence of linear-scaling diffusion Monte Carlo total energy per

atom to the value obtained with extended orbitals for bulk MgO. Open

squares with statistical error bars: the present method with cubic cut-

off. The dashed line shows the result with extended orbitals.

This shows that the energy appears to be converged within ~10

meV/atom for a cut-off radius of 6 au.


Organic and molecular magnetsOrganic and molecular magnetsS J Blundell and F L PrattJ. Phys.: Condens. Matter 16 R771-R828

Though magnetism is traditionally associated withmetals or simple compounds, some of the mostexciting recent studies in magnetism are on organicmaterials. Organic and molecular materials have anextraordinarily diverse range of magnetic properties.The use of organic and molecular building blockscan lead to new magnetic materials with intriguingproperties which have, on occasion, led to thediscovery of new physics. Stephen Blundell (Oxford)and Francis Pratt (Rutherford AppletonLaboratory)review this rapidly emerging field.

Purely organic ferromagnets, based upon nitronylnitroxide radicals, show long-range magnetic order at very low temperatures in the region of 1 K, whilesulfur-based radicals show weak ferromagnetism attemperatures up to 36 K. It is also possible to preparemolecule-based magnets in which transition metalions are used to provide the magnetic moment, butorganic groups mediate the interactions. Thisstrategy has produced magnetic materials with alarge variety of structures, including chains, layeredsystems and three-dimensional networks, some ofwhich show ordering at room temperature and someof which have very high coercivity. Even if long-range magnetic order is not achieved, the spincrossover effect may be observed, which hasimportant applications. Further magnetic materialsmay be obtained by constructing charge transfer salts,which can produce metallic molecular magnets.Another development is single-molecule magnets,formed by preparing small magnetic clusters. Thesematerials can show macroscopic quantum tunnellingof the magnetization and may have uses as memorydevices or in quantum computation applications.

The search is currently on for new materials withhigher coercivities, higher transition temperatures,more pronounced anisotropies, or largermagnetoresistance. Many puzzles remain to be solvedin, for example, the behaviour of single-moleculemagnets and the relationship between magnetismand superconductivity. Moreover, little has so farbeen done to understand how domains behave inorganic and molecular magnets. The advancesalready gained in this field have originated throughdevelopments in synthetic and coordinationchemistry, but the whole enterprise is aninterdisciplinary effort, involving both physicists andchemists, which promises to remain fast moving andcontinually surprising.

Coarse grain simulation of soft matterCoarse grain models and the computer simulation of soft materialsSteve O Nielsen, Carlos F Lopez,Goundla Srinivas and Michael L KleinJ. Phys.: Condens. Matter 16 R481-R512

Experimental work on complex condensed matterspans a broad range of temporal and spatial scales,from femtosecond dynamics and atomistic detail toreal-time macroscopic phenomena. Simulationmethods in which each atom is explicitly representedare well established but have difficulty addressingmany cooperative effects of experimental andtheoretical interest. Bridging the disparate time andspatial scales is possible with multiscale modeling, inwhich the various levels of treatment are coupled andfed back into one another.

Michael Klein and colleagues from University ofPennsylvania review a coarse grain (CG) simulationmethod that has ready access to events on these scales.They give some illustrative examples from biologyand materials science and they summarise existingsimulation techniques which access intermediatetimescales and length scales. They provide insightsinto the model building process and discussconsequences arising from the loss of detail. Theylook at a few situations in which theoreticalpredictions can be evaluated by CG simulationmethods, and at several situations in which the CG method goes beyond current theoretical and experimental reach. Finally they give someperspectives on future directions for research usingCG models.

Many topics are amenable to study with the CGmethod. Bilayers and monolayers involving a fewlipid species and cholesterol are suitable for studyingraft formation. The compression/expansion cycle ofthe lung surfactant DPPC could be studied in thepresence of the surfactant proteins (SPs), which areknown to alter monolayer collapse. Lipid mediatedprotein-protein interactions can be used to exploremembrane protein crystallization. The cyclic D,L-�-peptide nanotubes of Ghadiri et al could be studiedfor their self-assembly and membrane disruptionproperties. Monolayer structure at solid/waterinterfaces displays novel geometry such ashemicylindrical micelles which is being elucidatedwith atomic force microscopy. Entropic and enthalpicinteractions between amphiphile surfaces such asmicelles, lipid bilayers, microemulsion droplets, andcombinations thereof can be computed as potentialsof mean force. Protein alignment can be studied as afunction of surface pressure in Langmuirmonolayers. Self-assembled vesicles from non-lipidspecies such as surfactant-like peptides or blockcopolymers offer alternatives for many applicationsincluding targeted drug delivery. In conclusion, thereare clearly many possible future applications of CGmodels.

Simple models of protein foldingSimple models of protein folding and ofnon-conventional drug designR A Broglia, G Tiana and D ProvasiJ. Phys.: Condens. Matter 16 R111-R144

The protein folding problem, one of the greatunsolved problems of science, is simply stated: howdoes the one-dimensional structure of a protein (asequence of amino acids) determine the three-dimensional structure into which it folds? An evenmore elusive goal is the prediction of the catalyticactivity of an enzyme from its amino acid sequence.This is the subject of a review by G Tiana andcolleagues from Milan.

While all the information required for the folding ofa protein is contained in its amino acid sequence, it isnot yet known how to extract this information topredict the three-dimensional, biologically active,native conformation of a protein from its sequence.Using insights obtained from simple modelsimulations of the folding of proteins, in particularthe fact that this phenomenon is essentially controlledby conserved (native) contacts among (few) stronglyinteracting (‘hot’), as a rule hydrophobic, aminoacids, which also stabilize local elementary structures(LES, hidden, incipient secondary structures such as_-helices and _-sheets) formed early in the foldingprocess and leading to the postcritical folding nucleus(i.e. the minimum set of native contacts which bringsthe system beyond the highest free-energy barrierfound in the whole folding process) it is possible towork out a successful strategy for reading the nativestructure of designed proteins from a knowledge ofonly their amino acid sequence and of the contactenergies among the amino acids. Because LES haveundergone millions of years of evolution toselectively dock to their complementary structures,small peptides made out of the same amino acids asthe LES are expected to selectively attach to thenewly expressed (unfolded) protein and inhibit itsfolding, or to the native (fluctuating) nativeconformation and denature it. These peptides, ortheir mimetic molecules, can thus be used as effectivenon-conventional drugs to those already existing (anddirected at neutralizing the active site of enzymes),displaying the advantage of not suffering from theincrease in resistance.

See also a recent review by F Galisteo-González(Granada) and co-authors.

Measurement of interactions between proteinlayers adsorbed on silica by atomic forcemicroscopyJ J Valle-Delgado, J A Molina-Bolívar,F Galisteo-González, M J Gálvez-Ruiz,A Feiler and M W RutlandJ. Phys.: Condens. Matter 16 S2383-S2392

Using an atomic force microscope and the colloidprobe technique, they investigated the interactionforces between bovine serum albumin (BSA) layersand between apoferritin layers adsorbed on silicasurfaces.

The measurements were carried out in an aqueousmedium at different pH values and NaClconcentrations.


The atomic structure of a small protein, Chemotrypsin Inhibitor 2. The

dark grey curve highlights the chain structure.

The magnetic susceptibility of galvinoxyl. Ferromagnetic interactions

dominate above the phase transition at 85 K. The hightemperature

behaviour fits to a one-dimensional ferromagnetic Heisenberg model and

the low-temperature behaviour fits to a singlet-triplet model.

Self-assembly of a worm-like micelle as observed in CG simulations.

Simulations are started from a lattice configuration (A) with 100

EO40EE37 diblock copolymers in 20 000 CG water sites. Aggregation

of the hydrophobic blocks into two clusters (B), (C) is followed by their

merger (d) into a worm-like micelle which spans the simulation unit cell,

shown from two viewpoints (E), (F). The EO monomers are shaded

dark; the light EE monomers form the core of the micelle.

TOPICAL REVIEWS ANDSPECIAL ISSUE PAPERS


Self-assembled monolayersSelf-assembled monolayers: from ‘simple’ modelsystems to biofunctionalized interfacesFrank SchreiberJ. Phys.: Condens. Matter 16 R881-R900

Frank Schreiber (Oxford) reviews recentdevelopments in the area of self-assembledmonolayers (SAMs) and their applications. There is increasing interest in soft condensed matter, and in thin films additional issues related to the reduceddimensionality come into play.

The term ‘self-assembly’ may be defined as thespontaneous formation of complex hierarchicalstructures from pre-designed building blocks,typically involving multiple energy scales andmultiple degrees of freedom. Specifically, self-assembled monolayers are ordered molecularassemblies that are formed spontaneously by theadsorption of a surfactant with a specific affinity of itsheadgroup to a substrate. They are usually preparedfrom solution, although some systems can beprepared from the vapour as well.

The discovery of SAMs has transformed surfacechemistry and also led to new physics. It has broughttogether the study of well defined inorganic surfacesand organic species, which from a physics perspectivewere previously often considered rather undefined.The great flexibility of the concept of SAMs broughtabout by the wide choice of endgroups which can beanchored to the substrate has led to a broad range ofapplications of SAMs including importantdevelopments in the area of biotechnology.

Schreiber first discusses issues related to thestructure, the phase transitions, the phase diagram,and the growth dynamics. He explains how theinternal degrees of freedom and the multipleinteractions involved can lead to fairly rich phasebehaviour even for systems which are commonlyconsidered ‘simple’ model systems. Then he discussesselected problems for more complex SAM-basedsystems, including SAMs as substrates for growth,SAMs and molecular electronics, electrochemicalapplications, and ‘switchable’ SAMs, as well as theuse of SAMs for biofunctionalized surfaces andlateral structuring.

The fundamental questions of adsorption,structure, phases, and phase transitions have beenthoroughly studied in the past, but several issuesremain unresolved, probably reflecting the complexcompetition of multiple interactions and degrees offreedom, giving rise to various structures which areenergetically similar. Much present and future work isrelated to utilizing the various ways to modify andfunctionalize surfaces by SAMs, with biorelatedapplications being the most dynamic area. SinceSAMs are not so much a specific class of compounds,but rather a very flexible concept with virtuallyunlimited potential for applications, we expect thatthe area of SAMs will continue to thrive.

Quantum cascade structuresCoherent charge transport in semiconductorquantum cascade structuresMichael Woerner, Klaus Reimann and Thomas ElsaesserJ. Phys.: Condens. Matter 16 R25-R48

Quantum cascade structures have wide application in electrically driven semiconductor lasers working inthe mid- to far-infrared spectral range. Opticalamplification in such unipolar devices is based on apopulation inversion between quasi-two-dimensionalconduction subbands in coupled quantum wells. Thepopulation inversion in the active region is generatedby electrons tunnelling from an injector regionthrough a barrier into the upper laser subband and byultrafast extraction of these electrons out of the lowerlaser subband through a barrier into the next injectorregion. Such transport processes on ultrafasttimescales have been the subject of extensiveexperimental and theoretical work without reachinga clear physical picture of the microscopic electrondynamics.

Thomas Elsaesser of the Max Born Institut andcolleagues review a comprehensive experimentalstudy of electron transport in electrically drivenquantum cascade structures. They investigatedultrafast quantum transport from the injector into the upper laser subband by mid-infrared pump-probeexperiments directly monitoring the femtosecondsaturation and subsequent recovery of electricallyinduced optical gain. For low current densities, lowlattice temperatures and low pump pulse intensities,the charge transport is dominantly coherent, leadingto pronounced gain oscillations due to the coherentmotion of electron wavepackets. For higher currentdensities, lattice temperatures, or pump intensities,the gain recovery shows an additional incoherentcomponent, which essentially follows the pump-induced heating and subsequent cooling of thecarrier gas in the injector.

Even at the high electron densities present in aquantum cascade laser the coherence properties ofthe electron wavefunction play an important role forthe microscopic injection process. This process iscrucial for generating gain in quantum cascade lasersand represents a key step in the overall chargetransport through the device. These results stronglysupport the empirical finding that the design of thewavefunction overlap between the injector subbandsand the upper laser subband is essential for theperformance of a quantum cascade laser. Theoreticalcalculations that include both the quantum characterof transport and the decoherence caused by electron-electron scattering are still lacking and pose achallenge for the future.

Slow and fast light in solidsUltra-slow and superluminal light propagationin solids at room temperatureM S Bigelow, N N Lepeshkin and RW BoydJ. Phys.: Condens. Matter 16 R1321-R1340

Recent interest in the old problem of how a wavetravels through a dispersive material has beensparked by the discovery of systems that have highdispersion, yet allow a pulse to propagate relativelyundistorted. In addition, these new systems haverelatively low loss so the pulse dynamics are easy to observe. However, until now, all of the systemsdeveloped to generate slow or fast light have beendifficult to implement.

In this review, R W Boyd and colleagues atRochester University explore ways to produce slowand fast light in a room-temperature solid-statematerial. First they describe the concept of coherentpopulation oscillations—the primary physicalmechanism used to generate large dispersion. Theyshow that when the beat frequency between thepump and probe beams is slow enough, it will causethe population in a two-level atom to oscillate. Thistime-varying population will cause energy to bescattered out of the pump beam and into the probe,so the probe will see less absorption over a narrowfrequency range. Correspondingly, the group velocityfor the probe can be very large within the samefrequency range.

They describe their experimental demonstration of ultra-slow light propagation in ruby using coherentpopulation oscillations. They observed a groupvelocity as low as 58 m s-1. Their results included theobservation of a delay of both amplitudemodulations and pulses.

They showed how it is possible to observe bothultra-slow and superluminal group velocities inanother material, alexandrite. Since alexandrite is aninverse saturable absorber at certain wavelengths, thesign of the group velocity is changed. In alexandritethe chromium ions can occupy either mirror sites(having mirror symmetry) or inversion sites (havinginversion symmetry). As a result of the energy levelstructure at each site, ions at mirror sites experienceinverse-saturable absorption (fast light), whereas ionsat inversion sites experience saturable absorption(slow light). The competing effects from ions at eithersite can be easily distinguished because they havemarkedly different population relaxation times.

Finally, the authors discuss the significance of ultra-slow light propagation. While much work remains tobe done, they conclude that these slow lighttechniques could be very important in developing all-optical control of communication and storageapplications.


Schematic phase diagram of decanethiol on Au(111) in temperature

and coverage space. The different regions and phases are denoted as S

(stripes), IS (intermediate structures), C (c(4 �2)), and L (liquid).

The broken lines indicate phase boundaries of the IS, which are not

yet fully established. The solid curve between C and L (melting transi-

tion) exhibits a sharp rise near full coverage. Note that this is similar

to the behaviour found for much simpler systems such as nitrogen on

graphite

A conduction band diagram of the GaAs/Al0.33Ga0.67As quantum

cascade laser structure (sample A). Probability densities ��(z)|2 are

shown for the wavefunctions relevant for the QCL dynamics: |g� (ground

state in the injector), |3� (upper laser state), |2� (lower laser state) and

|1� are eigenstates of the electronic Hamiltonian without the tunnel

coupling through the injection barrier (the wide barrier to the left of the

active region).

The experimental setup used to observe slow light in ruby.



Read-out of single spinsRead-out of single spins by optical spectroscopyF Jelezko and J WrachtrupJ. Phys.: Condens. Matter 16 R1089-R1104

Quantum computing is tantalising. It exploits theidea that a quantum mechanical system can exist in asuperposition of several states at once, so it might beused for parallel data processing, enabling anexponential speed-up of certain calculations. Theopportunities are driving a growing interest inmanipulation and read-out of single-electron andnuclear spin states because of possible applications insolid state quantum computing.

F Jelezko and J Wrachtrup of University ofStuttgart review recent experiments on opticaldetection and manipulation of spin states of impuritycentres in a solid. They introduce the experimentalbackground of optical detection of single quantumsystems in solids. The interaction of an impuritycentre with an excitation field is analysed in terms ofoptical Bloch equations. They present recentexperiments on the electron spin resonance of singleorganic molecules and paramagnetic defect centresin diamonds and they show how the spin state of thenitrogen-vacancy paramagnetic defect in diamondcan be read out optically. They also discuss pulsedelectron spin resonance of the single paramagneticdefect centre in diamond.

The accurate measurement of a single spin statehas two important aspects. First, single-spin magneticresonance is a central point for any pure state basedquantum computing scheme. Several promisingtechniques are currently under investigation.Recently, controlled electron spin injection andsingle-spin detection were demonstrated usingelectrical read-out in quantum dots and scanningtunnelling microscopy of organic molecules.Important progress has been achieved in the field ofmagnetic resonance force microscopy, which recentlyshowed detection sensitivity of two electron spins. Yetoptical detection remains a unique technique,capable of demonstrating coherent ESR and NMRin experiments on single quantum systems. The nextstep will be to show coupling between several spins.This will allow achievement of two-qubit gates,which are basic elements for quantum computing.The second important field is of more fundamentalcharacter. Experiments with single spins are suitablefor experimental testing of quantum mechanics.Projective spin measurements on single quantumsystems can be used in tests of the quantum Zenoeffect and Bell’s inequalities.

Superconducting nanostructuresSuperconducting nanostructures fabricated with the scanning tunnelling microscopeJ G Rodrigo, H Suderow, S Vieira, E Basconesand F GuineaJ. Phys.: Condens. Matter 16 R1151-R1182

Low-dimensional conducting and superconductorsystems reveal much new physics, as is evident fromreports of experiments with single small particles,thin wires, carbon nanotubes or DNA molecules.Control of the nanoworld has been revolutionized bythe invention of the scanning tunnelling microscope(STM) and related techniques such as atomic forcemicroscopy, magnetic force microscopy and scanningHall probe microscopy.

S Vieira (Madrid and co-workers review theproperties of nanoscopic superconducting structuresfabricated with a scanning tunnelling microscope,with emphasis on the effects of high magnetic fields.These systems include the smallest superconductingjunctions which can be fabricated, and they are aunique laboratory in which to studysuperconductivity under extreme conditions. Thereview covers a variety of recent experimental resultson these systems, highlighting their unusual transportproperties, and theoretical models developed for theirunderstanding.

Charge transport through superconductingnanobridges can be dramatically modified by smallchanges in the minimal cross-section region, theneck, but the overall nanostructure (nanobridge)remains unmodified when scanning through theseregimes. At a high level of current, heating and othernonequilibrium effects appear. In atomic-sizecontacts superconductivity and quantum transportphenomena can be studied in a well controlledmanner. Breaking the tip into two parts results intotwo atomic size nanotips. One of these can be in situtransported elsewhere and used to perform atomicresolution microscopy and spectroscopy over asample, without change in vacuum or temperatureconditions. The application of an external magneticfield confines the condensate around the bridgeregion, creating a nanoscopic superconductor with aperfect interface with the normal region, solving in anatural way the contacting problems associated withthis kind of structure. This unique system gives us thepossibility to perform experiments in a highlycontrolled situation. Theoretical calculations usingGinzburg-Landau theory and Usadel equationsprovide a framework to understand the mostimportant aspects of superconductivity in thesebridges.

Publishing in Journal ofPhysics: Condensed Matterenables you to reach awide readership ofcondensed matter andmaterials physicists.

Readers and authors alike benefit from:● Rapid publication time● High editorial standards● Wide exposure through the electronic

journal● Large world-wide readership● Rising Impact Factor● Electronic submissions in any format● No page charges● A strong programme of invited topical

reviews and special issues● A complete online archive featuring our

award-winning electronic journals service● Quick and easy reference linking

How to submit your articleArticles may be submitted electronically viathe web www.iop.org/Journals/authorsubs orby e-mail [email protected]. If you are unable tosubmit electronically, please send three papercopies to:

Publishing Administrator, Journal of Physics:Condensed Matter, Institute of PhysicsPublishing, Dirac House, Temple Back,Bristol BS1 6BE, UK.

How to subscribe and contact informationFor more information on subscribing toJPCM or to request a free sample copy please visit the journal Webpage atwww.iop.org/journals/jpcm. Alternatively,please contact the relevant address below.

USA, CANADA & MEXICOInstitute of Physics PublishingThe Public Ledger Building, Suite 929,150 South Independence Mall West,Philadelphia, PA 19106, USA.Tel: (215) 627-0880 Fax: (215) 627-0879 E-mail: [email protected]

EUROPE AND REST OF WORLDCustomer Services Department,Institute of Physics PublishingDirac House, Temple Back, Bristol BS1 6BE, UK. Tel: +44 (0) 117 929 7481Fax: +44 (0) 117 929 4318E-mail@ [email protected]

The ODMR spectrum of a single pentacene molecule. The inset shows

the chemical structure of pentacene.

Sketch of the nanobridge fabrication process. Frames (a)-(f) illustrate

different stages of the process: (a) tip and sample in tunnelling regime;

(b) the tip is pressed against the substrate, both electrodes deform plas-

tically and forma connective neck; (c)-(e) indentation-retraction cycles

produce a plastic elongation of the neck; (f) the rupture of the nanobridge

takes place.

AUTHOR INFORMATION

Documents

Journalof Physics: Condensed Matter - UCLchianti.geol.ucl.ac.uk/~dario/pubblicazioni/jop_top_papers04.pdf · 2 CONDENSED MATTER: TOP PAPERS 2004 SHOWCASE Magnetism in carbon nanotubes