europh ysicsnews · 3page31 page23 page 18 number3• volume38 • 3 contents europh ysicsnews...

europhysicsnewsVolume38

•number

3

Institutionalsubscriptionprice:

99euros

peryear

38/32007

Condensed matter theory by computer simulationToying with physics

Fun with the setting sunBoundedmodes to the rescue of optical transmission

A thermo-magnetic wheel

European Physical Society

� PAGE 31

� PAGE 23

� PAGE 18

number 3 • volume 38 • 3

contents

europhysicsnews

europhysicsnews2007 • Volume 38 • number 3

Condensed matter

theory by computer…

Toying with physics

Fun with the setting sun

Cover picture: See Figure 2, p.21 in article “Condensed matter theory bycomputer simulation: from materials to chemical biology”

EDITORIAL05 My views and programme

Friedrich Wagner

07 Curiculum Vitae09 Laudation Ove Poulsen

Friedrich Wagner

NEWS10 Denis Jérome, Editor-in-Chief of EPL during an Eventful Time

ASSE 2006 (Advanced School on Space Environment)11 Prof. V. Dose, new Editor in Chief of EPL12 Ioan Ursu 1928 – 200714 India study tour 2006: experiencing physics abroad15 APFA5: Ties between Physics and Economics

HIGHLIGHTS15 Charge density waves in manganites16 Modelling the inner life of cell adhesion clusters

Sequential interatomic decay in argon trimers17 Plasma spraying for solid oxide fuel cells

A Mirror to generate a beam

FEATURES18 Condensed matter theory by computer simulation: from materials to chemical biology

Mauro Ferrario, Giovanni Ciccotti and Kurt Binder

23 Toying with physicsHassan Aref, Stefan Hutzler and Denis Weaire

27 Bounded modes to the rescue of optical transmissionMichaël Sarrazin and Jean-Pol Vigneron

31 Fun with the setting sunL.J.F. (Jo) Hermans

31 A thermo-magnetic wheelClaudio Palmy

My views and programme [EDITORIAL]


editorial

Volunteers are important actors in many of the activities of the European Phys-ical Society. The motives for their engagement vary though they are often

grounded in an appreciation of the physics field in which one is embedded, andfrom the insight that physics benefits if carried out jointly on a European scale.

In my own case it is rather obvious. The remarkable progress fusion researchhas made over the last decades, thanks to its European-wide organisation was adetermining factor in my involvement in the Plasma Physics Division as chair-man. The European dimension of plasma physics research allowed the buildingand operation of JET in the UK, currently the largest and most relevant fusiondevice, and it ultimately made Europe the best choice as the ITER site. ITER, tobe built in Cadarache, France, though still a scientific device, will be the first fu-sion reactor, producing a factor of 10 more power than externally invested.

Another reason for my European engagement stems from my German origin.My generation was brought up with the view that the German post-war politicalproblems would be solved by and within the European unification process andthat the future would be European. The European Union that was founded exactly50 years ago with the conclusion of the Treaty of Rome (March 1957) led to the cre-ation of both the EEC and Euratom. The fusion programme is part of the Euratomcontract. Science has profited from this integration but it has also contributed in aunique way to the evolution of a European identity: science creates values.

An exciting decade lies ahead for physics. Large and promising devices havebeen approved and will start operation in the near future. In this year and next,science in Europe will carry nature back to its beginnings and conditions will becreated with the LHC, the Large Hadron Collider at CERN, closer than ever tothose shortly after the Big Bang.

The 7th Framework Programme foresees a budget of more than 54 Bn € for re-search. The support of basic research with 7.5 Bn € and the creation of the Eu-ropean Research Council establishes a new horizon for science. The creation ofthe ERC will introduce an element of scientific self-organisation in the 7th Frame-work Programme and entrust scientists with more decision power in the sup-port of projects and the distribution of resources for research. Theimplementation of the Bologna process is expected to lead to a higher mobilityof students and will introduce competitive elements in attracting the best ofthem. The creation of the European research area will continue and, from themany steps already taken and those foreseen, it will gain further momentum.

The enthusiasm for these developments should not distract from reality. Thefunds made available by the European union for R&D account for only 6% of theEU overall budget for research. Many initiatives still suffer from the bureaucraticstructure of distant institutions and, for sure, every effort to make administrativeprocesses simpler will pay off. The same is true for making the applicationprocess more transparent and more predictable in order to prevent investmentsof time, motivation and resources by the applicant that are in vain.

Serious concerns remain about physics education in schools and its continu-ation in universities. I support the basic principles expressed in the EPS educa-tional policy position paper: Bachelor education in physics should have a largelycommon curriculum to facilitate student mobility between universities andcountries, both during and after the bachelor programme; Masters degrees needto be more varied and involve specialisation in particular areas of physics orcombinations of physics with related disciplines; and Doctoral degrees shouldnot be modularized to any large extent.

These principles also address major problems associated with the Bolognaprocess. The Bachelor in Physics may not meet the needs of industry unless cur-ricula are adapted. On the other hand, it may lead to a reduction in the numberof PhD students and weaken the scientific base. The major concern of the EPS •••

europhysicsnews2007 • Volume 38 • number 3

europhysicsnews is the magazine of the Europeanphysics community. It is owned by the EuropeanPhysical Society and produced in cooperation withEDP Sciences. The staff of EDP Sciences are involvedin the production of the magazine and are not re-sponsible for editorial content. Most contributors toeurophysicsnews are volunteers and their work isgreatly appreciated by the Editor and the EditorialAdvisory Board. europhysicsnews is also availableonline at: www.europhysicsnews.comGeneral instructions to authors can be found at:www.eps.org/publications.html

Editor: Claude SébenneEmail: [email protected]

Science Editor: George MorrisonEmail: [email protected]

Executive Editor: David LeeEmail: [email protected]

Graphic designer: Xavier de AraujoEmail: [email protected]

Director of Publication: Jean-Marc Quilbé

Editorial Advisory BoardChairman: George MorrisonMembers: Alexis Baratoff, Giorgio Benedek,Marc Besançon, Charles de Novion, Carlos Fiolhais,Bill Gelletly, Martin Huber, Frank Israel, Thomas Jung,Jean-Marc Quilbé, Christophe Rossel, Claude Sébenne,Wolfram von Oertzen

© European Physical Society and EDP Sciences

EDP SciencesManaging Director: Jean-Marc Quilbé

Production: Agnès HenriEmail: [email protected]

Advertising: Isaline BoulvenEmail: [email protected]

Address: EDP Sciences17 avenue du Hoggar, BP 112, PA de Courtabœuf,F-91944 Les Ulis Cedex A • Francetel: +33 169 18 75 75 / fax: +33 169 28 84 91web: www.edpsciences.org

EPS Secretariataddress: EPS • 6 rue des Frères LumièreBP 2136 • F-68060 Mulhouse Cedex • Francetel/fax:+33 389 329 440/449 • web:www.eps.org

The Secretariat is open 09.00–12.00 / 13.30–17.30 CETexcept weekends and French public holidays.

SubscriptionsIndividual Ordinary Members of the European PhysicalSociety receive europhysicsnews free of charge.Members of EPS National Member Societies receiveeurophysicsnews through their society, except membersof the Institute of Physics in the United Kingdom andthe German Physical Society who receive a bimonthlybulletin. The following are subscription prices availablethrough EDP Sciences.

Institutions 99 euros (VAT included, European Unioncountries); 99 euros (the rest of the world).

Individuals 58 euros (VAT included, European Unioncountries); 58 euros (the rest of the world).

Contact: [email protected] visit www.epdsciences.com

ISSN 0531-7479ISSN 1432-1092 (electronic edition)Printer Rotofrance • Lognes, FranceDépôt légal: Mars 2007

••• as expressed in its policy statement onPhD degrees is the strong and centralisedinclusion of the PhD thesis work into thegoal of European harmonisation ofhigher education. A large fraction of re-search is carried out by PhD studentsunder the guidance of their supervisors.To change the weighting in favour ofmore educational and less research ele-ments in the preparation for a PhD the-sis may lead to a set-back of Europeanresearch. The training PhD students re-ceive is not only a first step in an aca-demic research career but also for a joboutside academia. The practical treat-ment of a research problem during aPhD programme is the best preparationfor both career paths.

A major goal of EPS is to contributeto the development of balanced condi-tions for physics in all of Europe. Theconsequences of the termination of theSoviet Union represented a tremendouschallenge for the maintenance of thisgoal and principle. EPS has helpedphysics in the FSU countries to organisethemselves under democratic gover-nance and to establish their national rolesand to develop contacts within the Euro-pean research structures. This processhas not yet come to an end but the trendto normalisation is evident. The EPSGroup “Physics for Development” isstrongly motivated to help the countriessouth of the Mediterranean by selectededucational programmes, providing sup-port for summer schools and by givinggrants to enable contacts to be made withour research network. However, a longway is still ahead of them (and us).

A topic which deserves and requirescontinuous attention is to improve thechances of women in physics. The moti-vation may not necessarily be altruisticbut may well have its origin in the recog-nition that physics cannot do without50% of its talent and intellectual poten-tial. I hope everyone will agree when Isay that progress in physics needs alsothe perspective and insight of womenand that research teams benefit from fe-male scientists in their staff. Wheneverthere is a possibility, EPS should look fortalented women to fill a vacancy in itsown structures. Its policy must be thatthe Divisions and Groups welcome

women onto their Boards, that womenare invited to become members of con-ference Programme Committees andthat the outcome of these organisationalsteps bears fruit – that more women giveinvited talks at conferences in recogni-tion of their capabilities, and improvetheir chances to advance their careersand play a more accepted role. EPS willcontinue to support women in their in-terest to develop networks and to formand maintain organisations to promotetheir case. As the necessity to provideequal opportunities is also a task of na-tional societies, better cooperation andcoordination might accelerate thisprocess toward normalisation.

EPS encourages the development of aphysics publication platform made up ofscientists, learned society publishers, andcommercial publishers. The services sci-ences require from publishers of scien-tific journals are the maintenance of ahigh scientific standard by an efficientrefereeing system, electronic accessibilityin a form which fosters international co-operation and an interdisciplinary ap-proach, and good functionality fordocumentation and retrieval. Scientistsexpect from publishers and journals fairtreatment of their papers, quick publica-tions both on the internet and in print,services regarding the quality and acces-sibility of the paper as well as on-lineavailability of archived papers, and ajournal policy which continuously in-creases the impact factor and thus im-proves the reputation of its papers and ofits contributing authors.

Papers with the highest prestige in ourfield are mostly published in journalsbased in the United States. This is partlydue to the excellent research carried outby our American colleagues and friendsbut it is also owing to the circumstanceof a homogeneous research area withinthe US, with publication in one language.In Europe, our long research tradition re-flects the national history in science andthe publication landscape is more di-verse. Many national journals have pub-lished what is now to a large extent thebasis of our physics inheritance. It alsohas to be recognized that the nationalphysics journals of Eastern countrieshave played an important role after the

breakdown of the Soviet Union as focalpoints in the phase of restructuring.

The attractiveness of APS journalswould be acceptable if there were not aclose link between international recogni-tion of the journal (focused on its impactfactor) and career chances of the authors.The consequence is that the system has atendency to instability, to an increasingconcentration on a few highly attractivejournals. Understandably, authors tend tosubmit to the journal with the highestprestige. It has become a sportive elementin submitting papers first to the mostprestigious journal at the expense of pos-sibly losing some time in the case of itsrejection. Rejected papers are re-submit-ted to what develops as a second categoryof journals – at least in the eyes of thosewho decide on careers.

Dissemination of scientific results isa major task of EPS. We will, therefore,have to watch carefully the reaction andcountermeasures of European scienceand science managers to the trend that –to a large extent - scientific achievementsare not published and archived wherethe underlying research has been fi-nanced and the scientific results havebeen achieved. Fair and open competi-tion is a fundamental principle of theEuropeanisation process. The EPS isconcerned with the fragmentation of thepublication industry in Europe and itsconsequent contribution to the growingmonopoly of US publications. Scientistsneed more choice of high quality publi-cations. Europe’s publishers shouldtherefore be supported to improve theattractiveness of their journals. Thefoundation of an umbrella organisationof European based publishers wouldhelp and ease the dialogue betweenpublisher and the national or EU-widerepresentation of physics research. Sci-entists, on the other hand, should con-template the mechanisms which causethe publication market to concentrateand whether this – in the long run - is intheir true interest. Close cooperation be-tween scientific community and regionalpublishers is necessary to avoid a loss ofdiversity with consequences that are dif-ficult to foresee. EPS will carry out amajor campaign in 2007 to enhance thevisibility and attractiveness of European

6 • volume 38 • number 3 europhysicsnews

editorial


editorial

journals in general, with a specific focuson EPL.

Each president has to formulate “his(or her)” programme for his (or her)two-year term. As presidents come andgo it is mandatory to respect continuitye.g. in the topics mentioned above. Ofparticular importance to me is that theDivisions and Groups are encouragedand supported to carry out the tasks oftheir mandates, predominantly the or-ganisation of conferences and the dis-semination of scientific results. EPSshould continue to offer the most rele-vant conferences in the respective fieldsand to make the good ones even better.The scientific material presented in theseconferences is of high value and shouldnot be lost. In many cases the conferenceorganisers take care that the material ispublished as refereed papers. Where thisis not the case appropriate forms of pub-lication and archiving have to be found.Close collaboration with the EPS Secre-tariat is offered. The electronic means ofcommunication and exchange will con-

tinuously be improved. The work of theDivisions should also receive a higherawareness in Executive Committee andCouncil and the Divisions should be bet-ter integrated into the overall develop-ment of EPS. A forum has been initiatedwhere the Divisions and Groups of EPScan communicate with each other andwhere exchange with other EPS struc-tures is possible.

Another element I am interested in isto intensify the cooperation with the Na-tional Societies. Although national sup-port will continue to be their majorfinancial source, research and develop-ment will increasingly be shaped by theEuropean Commission and its organisa-tions. To make best use of this develop-ment, we should aim at closer cooperationbetween National Physical Societies andEPS. The agenda for the Council meetingshould therefore provide room also forjoint strategic discussions.We have to fol-low the political developments , the socialtrends of the EU on aspects of physics ed-ucation and research, to exchange the nec-

essary information and to analyse it as acommunity. We should also seek an un-derstanding of the broad spectrum of re-sponses within the largely differentnational societies.

This exchange of views and the jointplanning should be at all levels of the EPS:• The Council should promote exchange

of information between the national so-cieties. EPS and national physical soci-eties should identify joint projects. TheCouncil should be used to define struc-tures and distribute tasks to tackle andsolve them. An example would be thecontinuous analysis of the impact of ahomogeneous bachelor/master educa-tion in physics. In order to make thistremendous task possible, it might besufficient that each national societyidentifies one or two physics depart-ments where the information is pro-vided, exchanged and analysed. Thus, wemay obtain an early view on this devel-opment before the large and officialstudies appear, that will hardly single outand reflect the specific issues of physics. •••

Scientific CVI studied physics at the Technical Uni-versity in Munich (TUM) and I got myPhD degree in 1972. This period at theTUM was exciting because a newphysics department was founded withyoung professors, many returning fromthe States. My field was low-temperaturephysics, studying rotating superfluid Heand the question whether the surfaceexcitations of HeII were quantised likethose (rotons) of the bulk (they are not).As a post-doc at Ohio-State University,I expanded to superconducting metalsstudying the questions whether con-duction electrons contribute to the heattransfer from metals to Helium (Kaptizaresistance). Having experienced the firstenergy crisis in the USA, I happily ac-cepted an offer of the Max-Planck-Institute for Plasma Physics (IPP) inGarching/Munich. It implied a jump intemperature scale from K to eV (plasmanotation). I started as a diagnostician onthe Pulsator tokamak and continued on

the ASDEX tokamak being in charge ofthe physics of neutral beam heated plas-mas. Into this period fell the discoveryof the H-mode (the H-mode is a dy-namic state of the plasma with the in-teresting feature that stronger thermalforces lead to lower turbulent fluxes; thetransition into this regime happens in aself-organised form. The plasma in theH-mode is closer to the programmaticgoals of fusion). In 1987, I worked withthe Princeton team on their TFTR toka-mak. After 15 years of tokamak re-search, I felt it was time to change and Itook over the leadership of the experi-mental stellarator programme of my in-stitute. Now, I am more than 15 years inthis field.

The physics of high-temperature fu-sion plasmas has many interesting andrelevant features – aspects of magneticfield geometry, equilibrium, stability,collisional and turbulent transport,plasma wall interaction, heating and di-agnostics. During a long professional

life in this field (I was born in 1943), onegets in touch with all issues becauseplasmas are complex and involved,highly non-linear systems. My main in-terest is, however, plasma confinement.

Career CVI had a teaching assignment at theRuprecht-Karls University in Heidel-berg, was honorary professor at theTUM and I am at present full professorat the Ernst-Moritz-Arndt University inGreifswald.

In 1988, I was nominated Fellow ofthe Max-Planck-Society and Director atIPP. I was project head of ASDEX toka-mak, W7-AS stellarator and for 2 1/2years of the Wendelstein 7-X stellaratorproject to be built in the new MPG in-stitute in Greifswald right at the BalticSee. I am head of the experimental de-partment E3 since my nomination,served for 12 years in the Directoriumof IPP and was spokesperson of theGreifswald branch for 8 years.

Curiculum Vitae

• The divisions and groups should formcloser cooperation with their nationalcounterparts, wherever such structuresexist. EPS divisional conferences oftencooperate with the local institutions inthe organisation of conferences. This el-ement of cooperation and agreementshould represent a natural part of theEPS culture. Whenever a divisionalBoard meets in an EPS member country,an exchange with representatives of thenational science field should become anagenda item as a matter of course.

• Those who play a specific role in the so-cieties (EPS and national), because theydeal with administrative or legislativeissues or they guarantee the societies’continuity, have a specific and generalresponsibility to develop their own per-spective. They should also create formsof exchange and cooperation.

Another topic I am concerned with isenergy (actually, the concern of many).Europe has a need for the availability andaccessibility of a reliable flow of energyto maintain the productivity and com-petitiveness of its economy and the equi-librium of its societies. Europe’s ownenergy sources are running out and it re-lies increasingly on imports at a timewhen for the first time prices of the rawmaterials are set by market laws. The in-crease in gasoline prices provides us witha daily bulletin on the growing scarcityof this commodity.

One way of looking at the EPS is thatwe represent the knowledge of 100000physicists. Physics is playing a major rolein the energy issue through the develop-ment of fission energy, in providing nu-clear data for alternative processes, in thedevelopment of new concepts such as ac-celerator-driven sub-critical reactors, andin the storage, reprocessing and trans-mutation of waste. Physics also plays anobvious role in the development of fu-sion energy in the form of magnetic con-finement or of inertial fusion. But physicsis also of high importance in the tech-niques of renewable energy and in en-ergy-saving technology. Photovoltaicelectricity production is still a rich re-search field with many unexplored pos-sibilities using new materials, plasticsubstances, material combinations or theintegration of nano-technologies. Excit-

ing discoveries and technological break-throughs can be expected in this area.Similar arguments on the importance ofphysics apply to the development of fuelcells and of techniques to store energy -specifically electricity.

Energy is the topic of the EPS Tech-nology Group (TG). In addition, energyis one of the topics of the Nuclear Physicsand the Plasma Physics Divisions. It playsa role in the Condensed Matter PhysicsDivision (photovoltaic processes, mate-rials, etc.) and in the Quantum Elec-tronic Division (light sources). ThePhysics in Life-Sciences Division is thebridge to biotechnologies. The Environ-mental Division is knowledgeable on theatomic and molecular physical-chemicalprocesses in the atmosphere and dealsthus with the environmental conse-quences of energy conversion. These dif-ferent activities will be jointly discussedwithin a newly founded Energy WorkingGroup of the TG. The work will concen-trate on science and technology not onpolitical or social aspects. We should

demonstrate, however, that physics is notsynonymous with nuclear energy only;our goal should be the development ofCO2-free electricity production and, inthe long run, of a hydrogen-based trans-port system.

Environment and global warming areconsidered to be supra-national topics.This is not the case, however, with energywhere the debate is mostly restricted tothe national needs, opportunities and in-terests. This is not an appropriate associ-ation because energy is the primary andenvironment the secondary problem. Aglobal secondary problem can hardly besolved by a local approach to the primaryproblem. Europe still lacks a joint energypolicy; solutions are sought nation-wideand the strategies are widely different. Ef-forts are underway to change this. Physicsshould be ready both nationally and EU -wide when this issue is and remains offront-page concern. I would suggest thatthe national physical societies form en-ergy working groups, if this is not yetdone. We recently had a meeting of the


editorial


editorial

existing groups (Denmark, DPG, Finland,Hungary, IOP, Lithuania, SFP, and Swe-den) and we realised that EU-wide coop-eration is necessary and timely.

Let me summarize: I would wish tocontinue the efforts of past Presidents tomake the EPS more effective and toachieve a better balance within EPS be-tween East-West, North-South, male andfemale physicists, physics and society, andsmaller and larger physical societies. Ishould also wish for EPS to continue its

care for its members, its involvement inphysics education, and its efforts to main-tain a diversified publication scene. Lastbut not least, I would wish that EPSshould continue its presence in the Euro-pean science policy sphere, with its goal -the promotion of the case for physics.Allthis must be carried out in a transparentand comprehensible way, governed by theethical principles of science.

How to do all this? This brings meback to where I started - the role of the

volunteers, those who make a donationto the society, not in money but, of morevalue, in their time. We must cooperatein a creative atmosphere; ideas have to betreated like a gift and the maintenance ofmotivation is the uppermost imperative.We do not cooperate under an employ-ment contract but under a social contract- to make the best use of the investmentof our own time and that of others. �

Friedrich Wagner, EPS President

Dear Ove,

one president comes, another one goes.I want to thank you, Ove, on behalf of themembers of the Executive Committee, theSecretariat in Mulhouse and of all of uswho are assembled here, thank you foryour leadership and your contributions toEPS. I am sure, everyone in this room willhave the desire to shake your hand laterand thank you personally.

I will not be able to reflect all initia-tives you took and honour all of them.You picked up many important issues,initiated many important and urgentsteps with the consequence that I foundit difficult for me to identify additionalurgent issues to further shape and ex-pand and complement the portfolio ofEPS activities.You put education and the publicationissue onto the table and on the agenda;you considered the general issue ofphysics and society and you organised orparticipated or represented EPS in manyimportant meetings.You were fully awareof what EPS should do and could man-age and you developed your concept andfollowed it through. This may have to dowith your Danish inheritance, decidewhere to go and then strictly follow thecourse. Nevertheless, you presented yoursuggestions, had them discussed, allowedothers to influence them, and then a de-cision was taken.You represented EPS asan open, constructive and determined

You get also a present from me.As youknow, I very much honour everythingaround Max Planck. Here is his book the“Philosophy of Physics”. I found it in anantique shop in Florida – a good sign forthose who retire, but this is not the issuehere. It fits to the memorable momenthere to quote a few sentences from thisbook because they deal with suprana-tional aspect of physics and fit to EPS andthey address the relation between physicsand ethics, an element of our field whereyou are strongly engaged yourself:“If science is unable or unwilling to ex-

tend beyond the limits of the nation it isunworthy of the name of science; and inthis connection physics enjoys an advan-tage over other branches of science. No-body will dispute that the law of natureare the same in every country; so thatphysics I not compelled to establish its in-ternational validity (…). Ethics also issupranational, otherwise ethical relationscould not exist between the members ofdifferent nations.Here again physics takesup a strong position. Scientifically it isbased on the principles that it must con-tain no contradiction, which is terms ofethics implies honesty and truthfulness;and these qualities are valid for allcivilised nations and for all time; so thatthis scientific principle may claim to rankamong the first and most important ofvirtues.” �

personality. I am happy for my own pres-idency to have you still on board.

You probably all followed this nationalcompetition last year – no I do not meanthe soccer championship but somethingmuch more exciting, the selection of thehappiest people on earth and – low andbehold – the happiest people are theDanes. This happiness index will now fur-ther increase, because one of the Danes,Ove Poulsen, will get a present now andthis is always a source of happiness.

An important issue for you, Ove, wasalways to reach, what you called, the for-ward looking dimension. You educatedand trained individuals and committeesto eventually stop talking about the pastand finally come to and start to adopt theforward looking dimension. But, youyourself also took the necessary steps inthe right direction. We physicists aretrained to and have the natural desire todo everything under quantified condi-tions. In order to quantify, in the future,the necessary steps when the stage of theforward looking dimension has beenreached, you can use now this present -David actually had the idea. It is a stepcounter, a pedograph, it is also called anodograph, not an ovegraph – but pho-netically rather close. From now on, theforward looking dimension has a scale.(Maybe, we have a picture of the deviceto add or a photo when I handed it over)

Laudation Ove Poulsenat the occasion of the end of his EPS presidency by the new president, F. Wagner


newsand

views

The retirement of Denis Jérome asEditor-in-Chief of Europhysics Let-

ters, at the end of his three-year term, is anexcellent occasion to recall his great mer-its in advancing this journal. EPL - andphysics in general - have greatly benefittedfrom the scientific leadership he broughtto this foremost letters journal publishedin Europe.

Denis Jérome’s actions as Editor-in-Chief, all based on his scientific stature andexperience from earlier editorial functionswith the Journal de Physique and the Eu-ropean Physical Journal-B, have been in-valuable for bringing EPL closer to the aimof becoming a letters journal exploring thefrontiers of physics that is recognised notonly in Europe, but worldwide. In tunewith this development, and to steer clear ofbeing viewed as journal with an exclusive-ly European outlook, the Board of Direc-tors of EPL has now introduced the newdesignation ‘EPL’ rather than the explicit‘Europhysics Letters’.

Denis Jérome’s research activities are de-voted to organic conductors. His discov-ery, in the early 1980s, of superconductiv-ity in some organic materials was an eventwith worldwide impact on condensed-

matter physics. In 1985 he was given theHolweck Prize, an honour awarded annu-ally, alternately to a British physicist by theFrench Physical Society and to a Frenchphysicist by the U.K.’s Institute of Physics.Moreover, Denis was elected, during histerm as Editor-in-Chief of EPL, as a mem-ber of the highly selective French Acade-my of Sciences.

During his tenure, Denis Jérome suc-ceeded, as part of the regular renewal ofthe Editorial Board, in persuading manyleading scientists to collaborate as Co-edi-tors, at the same time keeping in mindboth an appropriate coverage of fields andan adequate balance of nationalities.

Besides this great effort and the ever-present, daily duties of an Editor-in-Chief,Denis had also to face two major transi-tions in the operations of EPL. The Edito-rial Office was moved from the former lo-cation of the EPS Secretariat in Petit-Lan-cy to the new headquarters of EPS in Mul-house. As part of this move, a youngerteam under Yoanne Sobieski replaced thewell-trained former team under Staff Ed-itor Edit Thomas. In a new ProductionContract for EPL, the Institute of PhysicsPublishing (IOPP) joined the previous

partners EPS, the French Edition Diffu-sion Presse Sciences (EDP Sciences) andthe Società di Fisica Italiana (SIF), andthus strengthened the journal. Through-out all these changes, Denis Jérome hasbeen a very active participant and trustedconsultant.

EPS, as the learned society responsiblefor the scientific quality of EPL, as well asall the owners of EPL 1, have many reasonsto be grateful to Denis Jérome for his out-standing stewardship of EPL over the pastthree years. �

Martin C.E. Huber and Claude Sébenne

Denis Jérome, Editor-in-Chief of EPL during an Eventful Time [PUBLICATION]

1 The Owners of EPL are the four Category-AMembers, the European Physical Society, theFrench Physical Society, the Italian Physical Socie-ty, and the Institute of Physics (UK); and the nineCategory-B Members: the Austrian Physical Socie-ty, the German Physical Society, the HungarianPhysical Society, the Institute "Ruder Boskovic", theNetherlands Physical Society, the Portuguese Phys-ical Society, the Pool of Scandinavian PhysicalSocieties (namely the Danish, Finnish, Icelandic,Norwegian and the Swedish Physical Societies), theSwiss Physical Society, and the Turkish PhysicalSociety; as well as two Associate Members (theInstitute "Josef Stefan" and the Spanish Royal Soci-ety of Physics).

ASSE 2006 (Advanced School on Space Environment) [CONFERENCE REPORT]

From the 10th to the 16th of September2006, an Advanced School on Space

Environment (ASSE2006) was organ-ized jointly by Consorzio Area diRicerca in Astrogeofisica – CARA, Inter-national School of Space Science – ISSS(www.cifs-isss.org) and World Institutefor Space Environment Research –WISER (www.cea.inpe.br/wiser). Theschool was held at the Scuola SuperioreG. Reiss Romoli (SSGRR), an excellentHotel/School located in the medievaltown of L’Aquila, 100 km east of Rome,Italy, surrounded by the beautiful moun-tains of the Gran Sasso d’Italia.

The main goal of the school was to pro-vide an overall view of the main physical

processes acting on the sun-earth environ-ment treated as a unique system. This kindof approach allows a significant improve-ment in the global knowledge of the entiresystem, paying particular attention tochanges on the Sun, which influence thenear-Earth space and to the multiple phys-ical processes that occur sequentially ineach domain. It is a dramatic fact that ourunderstanding of the physical phenomenathat contribute to Space Weather is still toolimited to allow for reliable and useful fore-casting. Additional theoretical and dataanalysis efforts are strongly needed andyoung scientists are encouraged to contin-ue in this direction. Educating students to-day about this kind of problem assumes a

remarkable importance because of associ-ated economical and social repercussionsin the future.

Keeping this in mind, lectures were or-ganized in such a way as to allow the youngresearchers attending the school to becomefully aware of the interdisciplinary natureof this subject; moreover, their attentionwas addressed to the improvements andprogress which are needed for a better un-derstanding of the Sun-Earth domain as awhole, and to the critical aspects of SpaceWeather. Thus, lectures ranged from thesolar interior to the solar atmosphere, tointerplanetary space, magnetosphere, ion-osphere, geomagnetic storms and sub-storms. Moreover, the programme also •••

� Group picture of the attendeesof ASSE2006 school. Moving fromleft to right, arrows indicate thedirectors of the school: Y. Kamide,U. Villante, R. Bruno and A. Chian.

••• treated basic concepts such as non-linearprocesses in space physics, and magneticreconnection.

The school was organized in plenarysessions; there were 18 lecturers who gavea 80 min lecture each. Most of the invitedlecturers of ASSE 2006 derived from thelist of authors of the Handbook of Solar-Terrestrial Environment edited by Y.Kamide and A. C.-L. Chian, to be pub-lished by Springer in 2007. About 40 stu-dents with graduate and undergraduatebackground from 10 different countries at-tended the school; they were also given thepossibility to present their research activi-ty during the oral and poster sessions ded-icated to them throughout the week. The

directors of the course were:R. Bruno (IFSI-INAF, Rome,Italy), U. Villante (Universi-ty of L’Aquila, Italy),A. Chi-an (INPE, Sao Jose dosCampos, Brasil), Y. Kamide(STELAB, Nagoya Universi-ty, Japan). �

Roberto Bruno (Rome)

Professor Volker Dose, an active mem-ber of the physics community for 40

years, has been appointed Editor-in-Chief of Europhysics Letters (EPL;www.epljournal.org). Professor Dose willtake on the role of Editor-in-Chief fromProfessor Denis Jérome (Université Paris-Sud, Orsay) at the next meeting of theEPL Editorial Board to take place in Paris,11- 12 May 2007.

Following his appointment, ProfessorDose said, “This is an exciting time forEPL. In 2007, the journal starts the nextphase of its life: a new vision as a truly in-ternational journal at the forefront ofphysics, a new website, and new journalcover and format. I look forward toworking with the Editorial Board and theinternational scientific community togrow our journal.”

Professor Volker Dose took his PhD atthe University of Zurich in 1967 and in1971 was appointed Professor at the Uni-versity of Würzburg. Later, he became Di-rector and Scientific Fellow at theMax-Planck-Institute of Plasma Physics inGarching,and since 1991 has also been FullProfessor at the University of Bayreuth.Hisresearch experience is diverse, including

magnetism, surface and plasma physics,collisions and statistics, specificallyBayesian statistics. He has applied themethod of Bayesian statistics to largely dif-ferent problems, which range from X-rayastronomy to plasma physics and climateresearch.This breadth in his research is re-flected by the citation of the Robert-Wichard-Pohl-Prize 2005, which VolkerDose received from the German PhysicalSociety. It read: …for his outstanding inter-disciplinary contributions to the physics ofatomic andmolecular hydrogen, to the elec-tron band structure of solid states and to theBayesian probability theory.

Prof. Dose was Deputy Dean of theFaculty of Physics and Astronomy of theUniversity of Würzburg from 1976 to1978 and again from 1982 to 1983; from1983 to 1985, he was its Dean. From 1984to 1986 he was Chairman of the SurfacePhysics Division of the German PhysicalSociety and from 1985 to 1988 DeputyChairman of the German Vaccum Soci-ety; from 1987 to 1998 he was Chairmanof the Bavarian Section of the GermanPhysical Society and from 1987 to 2000,V. Dose was Co-Editor of AppliedPhysics A.

Volker Dose`s scientific achievementsare honoured by many awards. TheRobert-Wichard-Pohl-Prize has alreadybeen mentioned. He has also received theGrünewald-Medal of the city ofWürzburg and the Honorary PromotionDr. rer. nat. by the Faculty of Physics ofthe University of Kassel. �

news and views

Prof. V. Dose, new Editor in Chief of EPL [LAUDATION]

number 3 • volume 38 • 11europhysicsnews

Ioan URSU, the fourth President of theEuropean Physical Society, has passed

away. He suffered a fatal heart attack onApril 17 2007, aged 79.

The academic community will re-member Professor Ursu for his contribu-tions to physics, as a professor andresearcher in Cluj and Bucharest, as theDirector of the Institute of AtomicPhysics in Bucharest from 1969 to 1977,as the President of the National Com-mittee for Nuclear Energy from 1969 to1977, as the Prime Vice President of theNational Council for Science and Tech-nology and a member of the RomanianAcademy of Sciences. He has also had anoutstanding international profile.

Born in the Transylvanian villageManastireni, he began his studies innearby Cluj-Napoca. During the Sec-ond World War, he attended high schoolin Turda, as Cluj had been ceded toHungary, following the Vienna dictate.After the war he graduated from theFaculty of Physics and Mathematics atBabes-Bolyai University in Cluj, wherehe obtained his doctorate in 1956 andDoctor Habilus in 1967. He had a suc-cessful career in Cluj beginning as a lab-oratory demonstrator (preparator), andending as dean of the physics facultyand Vice-Rector of the University. In aprivate discussion he praised with mod-esty many of his university mentorsand colleagues, Augustin Maior, AurelIonescu, Ion Dragan, Victor Mercea andAurel Novacu.

In Cluj Professor Ursu became inter-ested with what would prove to be one ofthe great atomic physics ideas – electronicparamagnetic resonance. This led to afruitful international collaboration whichbrought him in contact with two excep-tional personalities: Freeman Dyson andSerg Turkevich. Two books published in1959 and 1965 have become textbooksfor generations of students. A grandiosevolume, “La résonance ParamagnétiqueElectronique”, published by Dunod-Paris,also appeared in 1968. In its introductory

presentation Nobel laureate Alfred Kastlerdeclares“Le Professeur Ioan Ursu de l’U-niversité de Cluj (Roumanie), est unexpert internationalement connu dansle domaine de la résonance parama-gnétique électronique”. In the Russiantranslation another Nobel laureate A.N.Prokhorov, expressed similar apprecia-tion. Further working visits in Germany,the United Kingdom, Switzerland, Italy,Russia, increased his reputation andrecognition in the scientific world. In Ro-mania he created the school of scientiststhat in Bucharest, Cluj and Iassy continuesuccessful research in electronic and nu-clear resonance.

His reputation led to his nomination asDirector of the Institute of Atomic Physicsin 1969, twenty years after it was foundedby Horia Hulubei. He also assumed im-portant state positions such as the presi-dency of the National Committee forNuclear Energy, and of the NationalCouncil for Science and Technology. Inthis capacity,unavoidably,he also assumedvarious high political tasks.

The next seven years were good yearsfor science and particularly for physics.The old institute quadrupled, buildingmany new basic facilities: a tandem, amodern centre for radioisotope produc-tion, a nuclear medicine centre, a newlaser building, a Mossbauer laboratory, acomputing centre. Magurele, the site ofthe institute, became also a universitycampus, with a modern building for thefaculty of physics, a lyceum, comfortablestudent hostels and a commercial centre.The economic situation of Romania atthat time was healthy enough to allowthis development. Moreover, in less than10 years, the scientific output in physicsreached almost two percent of the worldoutput which created an optimistic ex-pectation and attracted young scientists.Ioan Ursu played a decisive and vision-ary role in this process. In a speech in thefall of 1970, he described in detail the fu-ture institute – The Platform of Magurele- that he, together with his predecessor

Hulubei, had contemplated. His nationaland international status offered a guar-antee to the state leaders to allocate fundsand the work force necessary to imple-ment these plans. In a way this coincidedalso with the ambitions and interests ofthe Ceaucescu presidential couple. Theonly part of the plan that was not realiseduntil 1977 were of houses and villas thatshould have lined the road connectingBucharest to Magurele. This was toomuch for the spartan future reserved bythe party leaders for the population.Andno exception was admitted.

Professor Ursu never left his callingfor research and writing, authoring over200 research papers published in presti-gious journals and several books written[see References] in collaboration withRomanian and Russian scientists.

His national and international pres-tige increased steadily. He was elected amember of the Romanian Academy in1974. An impressive number of nationaland international societies offered himmembership. He was a member of theInternational Committee of Ampere As-sociation, of the Dubna Scientific Com-mittee, the Soviet Union Academy, andGovernor (1972 – 1973) and Member ofthe AIEA Scientific Committee amongother affiliations.

In recognition of his international sci-entific profile, he was elected as President


newsand

views

Ioan Ursu 1928 – 2007 [OBITUARy]

Alexandru Calboreanu,President Romanian Physical Society

of the European Physical Society (1976 –1978). Two large conferences were or-ganised in Bucharest: the EPS-GeneralConference “Physics and Energy” (1975)and the AMPERE Congress (1982) withthe participation of six Nobel laureates.These actions have had a beneficial im-pact for the visibility and prestige ofphysics and science in Romania. This canbe illustrated by the numerous prizeswon by the Romanian teams at variousphysics Olympiads.

Ioan Ursu was equally concerned withall problems that confronted the economysuch as energy and high technology. Morethan two dozen research institutes werecreated under his guidance. He was in-volved directly in elaborating the nationalnuclear strategy and in the selection of thepath to nuclear energy. The choice of theCanadian CANDU power plant proved tobe a very wise and profitable one. He wasdirectly involved in various technologicalaspects of producing heavy water indus-trially, in the methodology of preparingthe reactor fuel and in safety control. Anew institute for nuclear technology, anatural uranium fuel processing plant anda heavy water production unit startedoperation by the end of the 80’s. He wasinterested in establishing a theoreticalmethodology for isotopic separationbased on paramagnetic electron reso-nance. However this was the time whendark clouds covered Romania which soonreached the brink of collapse. Impover-ished by increasing debts and by the lowproductivity, the economy could no longersustain itself. A regime of drastic restric-tions accentuated the crisis that culmi-nated with the events of December 1989and the fall of communism.

Ioan Ursu has already lost his positionas director of the Institute of AtomicPhysics in 1976.The institute has been splitinto five components under an umbrellacalled the Central Institute of Physics. IoanUrsu retained the position of Vice-Presi-dent of the National Council for Scienceand Technology as well as other academic

and political positions. It was in this ca-pacity that he navigated through the re-straints of political, economical, andideological bounds to help people and in-stitutions to survive and even, in somecases, to prosper. He was very attentive tothe needs of anyone who asked for his sup-port. He tried to keep the contacts openoutside the country,both East andWest,al-lowing many people to travel abroad. It isamazing how in the midst of the harshestyears of the mid 90’s,he was able to initiateand support the Balkan Physical Union,becoming its first president in 1986. TheUnion played an extraordinary role instrengthening the contacts and collabora-tion between the physicists in the regionduring the next decade. In spite of the na-tional, religious and political diversity ofthe area,even in spite of the wars that haveshaken a part of it, physics has proved tobe an excellent mediator and factor of sta-bility through the rules of logic, fairnessand scientific interest.

After the turmoil of 1989, ProfessorUrsu went through a period of disap-pointment and uncertainty. An appealfrom EPS and a visit of the SecretaryGeneral in the spring of 1990 helped hissituation a lot though the press was notalways very friendly.

Professor Ursu remained close tophysics after 1989, helping many peopleeven when his official possibilities werereduced to nil. He remained interested inmany areas of physics, including nuclearenergy, then laser physics and of courseparamagnetic electronic and nuclear res-onance. He was very proud to hear thatthe Cernavoda CANDU power plant wasconnected to the grid although others areclaiming the credit.

Professor Ursu was an extrovert. Heliked people and approached them withthe openness and naiveté of the Transyl-vanian country man. He was popular andhe liked young people whom he encour-aged to take up responsibilities. When Iasked him about his greatest achieve-ment, he responded promptly that it was

his family, his wife Maria Lucia, and thethree children, Horia, Ionut and Ioana,two medical doctors and a physicist.

Many people will not forget him. TheBalkan Physical Union, at the GeneralConference in Istanbul (August 2006),paid homage to his entire activity androle in BPU. One year earlier, in Con-stanta, he was awarded the title of DoctorHonoris Causa.

It remains for further generations ofphysicists and historians of science toweigh Professor’s Ursu merits from theperspective of his country’s development.So far the young scientists, aware or not,make use of what Professor Ursu andmany people of his generation strove toimplement. This is in fact the naturalflow of life.

Professor Ursu passed away quietlyand contented that he had fulfilled hisduty on earth. �

References“La résonance paramagnetique électronique”,

Dunod-Paris, 1968

“Laser radiation interaction with metals”, byA. M. Prokhorov V. I. Konv, I. Ursu, I. N.Mihailescu; NAUCA (Moscow), AMA-ZON (USA&UK) and Editura Academiei

“Magnetic Resonance in Uranium Compounds”,(in Russian), ENERGYA Pub. House - 1982

“Nuclear Materials”, Pergamon-1985

“Physics and Technology of Nuclear Materials”,(in Russian), ENERGOIZDAT - 1988

europhysicsnews number 3 • volume 38 • 13

news and views

On 2 May 2007, Maurice Jacob, whowas President of EPS (1991-1993)died from a heart attack. The EPScommunity is very sad and expressesits deep sympathy to his family andhis friends, at CERN and elsewhere.An homage to Maurice Jacob will bepublished in the next issue of Euro-physics News

M. Jacob 1933 – 2007


India study tour 2006: experiencing physics abroad [EDUCATION]

Crispijn Jansen, Geert-Jan Besjes and Koen van den Dungen,Radboud, University Nijmegen, The Netherlands

Every October, ‘Marie Curie’, the stu-dents' association for physics and

astronomy in Nijmegen (The Nether-lands), organises a two-week study tour.The aim of these tours is to give studentsan idea of how physics works 'in real life'and, more importantly, in different cultur-al environments. This year, nineteenstudents and one associate professor wentto India and travelled from Bangalore, viaGoa, to Mumbai.

The scientific side of our programmecan be divided into two parts: visits toIndia's major research institutes and vis-its to smaller universities. The former at-tract most of the (inter)national funding,enabling them to compete with high-profile European institutes, whilst the lat-ter, lacking the funds for such a level ofresearch, were noteworthy for the oppor-tunity to interact with the local students.The two largest institutes we visited werethe Indian Institute of Science (IISc) inBangalore and the Tata Institute of Fun-damental Research (TIFR) in Mumbai.At IISc, we first were given several lec-tures by staff members of the Depart-ment of Astronomy & Astrophysics,giving us new insights into the latest as-tronomical research. The head of this de-partment, Prof. Arnab Rai Choudhuri,was already familiar to us, being the au-thor of a renowned textbook on plasmaphysics. After lunch we also visited thelaboratories, including the NMR Re-search Centre.

The largest research institute in Indiais TIFR. Their Department of Astronomy& Astrophysics operates the Giant Me-terwave Radio Telescope in Pune as wellas a Balloon Facility at Hyderabad to de-tect cosmic radiation. The Department ofNuclear & Atomic Physics employs bothpowerful lasers and heavy ion accelera-tors to simulate extreme nuclear situa-tions in their laboratories.

Speaking with Indian students is oneof the best ways to really learn aboutphysics in an entirely different environ-ment. During our visit to the Universityof Mysore we passed a cosy classroomwhere students gathered for a lecture. In-terested in experiencing their class, weasked our guide whether it would be pos-sible to join the Indian students. Theteacher was enthused by our presenceand a lively exchange of names, hobbies,experiences and lectures made for somevery fruitful interaction.After our super-visor had told the group which courseshe gives and how he teaches them, thelocal teacher followed his example andtaught us the Indian approach to Statisti-cal Mechanics.

A different, yet equally valuable, meet-ing we experienced at the University ofMangalore. The head of the departmenthad prepared an interesting formal gath-ering with students and staff from theiruniversity and our relatively small group.Even more interesting was the informal

tea-time that followed the formal gather-ing, where we discussed religion, mar-riage, exams and, of course, physics withthe Indian students. Unfortunately thefaculty had imposed a curfew, because ofrecent violent actions of some local ex-tremists. We foreigners were hardly in anydanger, but the locals were kept safe onthe campus. Thus, the evening was spentwithout their lively and lovely company.

Of course, physics isn't the only goodthing in life; we simply had to enjoy thestriking Indian landscapes, colourful citycentres and beautiful cultural sites. Indiashowed us nature's diversity from lushjungles (with its mosquitoes to keep uscompany) to impressive plains. One ofthe cultural highlights was a Hare Kr-ishna temple in Bangalore: enormous,richly decorated and filled with thesinging of its many devotees. Other ex-citing sites include a 40-meter Hindustatue facing the ocean, the beautifulPalace of Mysore and Mumbai's Gatewayof India.

In typically Indian fashion, ourscheduled flight home was non-existent.Nevertheless, late-night friendly servicebrought us home with only a few hoursdelay, although Air France did tem-porarily lose half our luggage in Paris.Exhausted and satisfied with perhapstoo many impressions, our tour, as allgood things in life eventually do, cameto an end. �

newsand

views

� The gateway in Mumbai

� The EPN group in Mangalore


news and views

APFA5: Ties between Physics and Economics [CONFERENCE REPORT]

Anna Carbone, (Torino)

The 5th International Conference on‘Applications of Physics in Financial

Analysis’(APFA 5) was held in Torino,fromJune 29 to July 1 2006. The Conferencewas chaired by Anna Carbone, PhysicsDepartment, Politecnico di Torino, Italy.The program consisted of 30 Plenary Lec-tures, 42 Oral presentations in 2 parallelsessions and 45 poster presentations. TheConference was attended by more than 180delegates coming from 30 countries.

The APFA conference series started in1999 to serve the increasing need tostrengthen existing links and build newties between the physics and economicscommunities. It is increasingly recognisedby both the communities that a numberof conceptual and methodological ap-proaches based on statistical mechanicsmay be employed to understand eco-nomic phenomena. In addition to tradi-tional economic notions of coordination

via the price system and strategic interac-tion, models of collective phenomena arenow making their appearance in manybranches of microeconomics. Other in-stances of the complexity approaches, fa-miliar to physicists appear in evolutionarygame theory, demand theory, behaviouraleconomics and social economics. Theanalysis of social phenomena,with the in-troduction of concepts such as scalingand universality, is leading to radicallynew insights and questions, both theoret-ical and empirical. All these approachesemploy analytical and numerical toolsfrom what is known as the science ofcomplexity, an interdisciplinary approach,initially used for the analysis of systemswith strongly interacting subunits inphysics and engineering.

The publication of the two issues “Ap-plications of Physics in Finance” (PhysicaA, forthcoming) and “Physics in Society”

(The European Physical Journal B, forth-coming) with a collection of 55 paperspresented at the Conference is a major re-sult for the advancement and divulgationof the results achieved in the latest yearsin this interdisciplinary area.

The next edition of the conference will beorganized in Lisbon, Portugal, from July 4th

to 7th (2007) by the Department of Quanti-tative Methods,IBS - ISCTE Business School.

Further details and the full-length pa-pers can be found at the websitewww.polito.it/apfa5. �

Analysis of heat capacity measurements(see Figure) on three manganite com-pounds, La0.5Ca0.5MnO3, La0.48Ca0.52MnO3

and Pr0.48Ca0.52MnO3 has established thatthe transition at which the stripe phaseappears in manganites is second order,and is accurately modelled as a Peierlstransition in a material which is disor-dered or contains impurities. This is thebehaviour expected of a charge densitywave in a disordered system, in contrastto the previously accepted picture ofcharges localised at atomic sites.

Previous studies had concluded that theonset of the stripe phase was associatedwith the simultaneous onset of ferromag-netism in La0.5Ca0.5MnO3. We have foundthat the transitions in materials with andwithout a ferromagnetic signal share acommon origin, since the entropy changeat the transition is very similar for all three

compounds.However, the origin is not as-sociated primarily with magnetic order be-cause, while the magnetisation varies by afactor of 100 between the compounds, thechange in entropy remains roughly con-stant. This overturns the traditional view,and suggests instead that the transition inLa0.5Ca0.5MnO3 is driven mainly by aFermi-surface instability. Our result addsto the growing body of evidence that the

stripe and ferromagnetic phases are muchmore similar than previously thought,andindicates that the colossal magnetoresis-tance present in manganites is driven bythe delicate balance between the ferro-magnetic and stripe phase.

These results call for a re-evaluationof the many strongly correlated electronsystems which exhibit stripe andcheckerboard phases (e.g. cuprates, nick-elates and cobaltites) in terms of a disor-dered charge density wave model whichproduces insulating behaviour withoutthe need to invoke strong electron-phonon coupling. �

S. Cox, J.C. Lashley, E. Rosten, J. Singleton,A.J. Williams and P.B. Littlewood,“Evidence for the charge-density-wavenature of the stripe phase in manganites”,J. Phys.: Condens.Matter, 19, 192201 (2007)

Highlights from european journalsCharge density waves in manganites

Adhesion of biological cells to externalsurfaces is the fundamental process thatensures that multicellular organisms keeptheir structural identity over time. More-over cell adhesion is crucial for manyphysiological processes, including woundhealing and the immune response. A cellsettles on a surface in a sequence of stepsin which the distance between cell andsurface gradually decreases from micronsto nanometer over the time course ofminutes. This process is active in thesense that a cell commits itself to firm ad-hesion only if it receives suitable signalswhen establishing adhesion.

During recent years, it has becomeclear that these favourable signals also in-clude mechanical cues, which are receivedthrough small clusters of appropriate ad-hesion molecules at the cell-substrate in-terface. These adhesion molecules appearto be pre-clustered with an average clus-ter size of three to four molecules, corre-sponding to a lateral size of nanometres,and then might grow to sizes of 10.000molecules in a micron-sized cluster, socalled “focal adhesions”.

The binding dynamics inside focal ad-hesions is known to be very fast, al-though it is very difficult to observe itexperimentally, due to the 200 nm reso-lution of optical microscopy. In a recentpaper, Thorsten Erdmann and UlrichSchwarz therefore studied the stochasticdynamics of adhesion clusters of varyingsizes in a theoretical model, which isbased on well established principles ofthe physics of cell adhesion.

The two researchers found that forsufficiently small clusters and cell-sub-strate distances, the adhesion dynamicsis characterised by rapid switching be-tween an unbound and a bound state ofthe adhesion cluster. Such “bistability”has been discussed before mainly in thecontext of biochemical decision-making,e.g. in cells that decide to divide into twodaughter cells.

In contrast, the bistability found inthis new study is based on physical bind-ing-unbinding processes and propertiesof the elastic cellular material. It mightallow cells to explore new surfaces by es-tablishing many small and very dynamic

adhesion sites. The model also suggeststhat upon encountering favourable con-ditions, such transient adhesions couldquickly be made permanent by small in-creases in size and a concomitant reduc-tion in the distance between cell andsurface. �

T. Erdmann and U.S. Schwarz, “Impactof receptor-ligand distance on adhesioncluster stability”, Eur. Phys. J. E 22, 123 (2007)


highlights

Modelling the inner life of cell adhesion clusters

We have investigated sequential inter-atomic decay in the argon trimer Ar3.Large Van-der-Waals clusters that ab-sorbed X-rays predominantly fragmentinto singly charged ions and neutrals:highly charged ions are hardly produced.Usually,Auger decay in clusters occurs inthe atom that absorbed a photon and

thus a highly charged atomic ion is firstproduced in the cluster. How does thishighly charged atomic ion transfer itscharge and energy to surrounding atomsin the cluster? In order to clarify thecharge and energy transfer mechanism,we have investigated the decay from theargon trimer in the 2p hole state, usingelectron-ion-ion-ion coincidence tech-nique. The relationship of the electronenergy and the kinetic energy release(KER) of the triply-charged argon trimerAr3

3+ in the fragmentation into Ar+-Ar+-Ar+ is shown in the Figure. The distribu-tion of the sum of electron kinetic energyand KER in Figure (d) shows the peak at13.6 eV.We can assign this peak to the in-teratomic Coulombic decay (ICD)process following the Auger decay.

The charge and energy transfer mech-anism in the argon trimer is found to beas follows. The first-step Auger decay ofthe 2p hole state in Ar3 leads to the one-site two-hole state Ar++-Ar-Ar that cou-

ples to the two-site satellite state Ar+*-Ar+-Ar. These states are subject to ICD tothe state Ar+-Ar+-Ar+ via the two-sitesatellite character. This second-step ICDprocess has been identified unambigu-ously by the electron-ion-ion-ion coinci-dence spectroscopy in which the kineticenergy of the slow ICD electron and theKER among the three Ar+ ions are meas-ured in coincidence.We believe that suchsequential interatomic processes thatlead to many singly-charged ions can beseen in all clusters, molecules, liquids andsolids, especially in biological moleculesin living cells. �

X.-J. Liu, N. Saito, H. Fukuzawa, y.Morishita, S. Stoychev, A. Kuleff, I.H.Suzuki, y. Tamenori, R. Richter, G.Prümper and K. Ueda, “Evidence ofsequential interatomic decay in argontrimers obtained by electron–triple-ioncoincidence spectroscopy”, J. Phys. B: At.Mol. Opt. Phys. 40 F1 (2007)

Sequential interatomic decay in argon trimers

� This shows a cell (orange, with nucleus) adher-ing to a substrate (blue, like glass) through twoadhesion clusters, each being composed of anumber of adhesion bonds, each of which can beeither closed or open. It is exactly the stochasticdynamics of opening and closing of these adhe-sion bonds which is modelled in our work. Thestructure in red is the actin cytoskeleton whichlocalizes the adhesion bonds into adhesion clus-ters and also puts them under mechanical stress.


highlights

The growing energy demand for the set-tlements of population needs to be bal-anced by the development of cost-effectiveand environmentally friendly technologiesof energy production. Solid Oxide FuelCells (SOFC) have an exceptional poten-tial for use as electric power generationsystems since they are fed, at high temper-ature (~900°C),with oxygen and hydrogengases to provide electricity and steamcleanly with a high energy efficiency.How-ever, much interdisciplinary research isstill carried out that is focused especiallyon the reduction of the operating temper-ature of SOFC which leads to a decrease inthe cell performance. Research efforts arecurrently devoted, on the one hand, to thedevelopment of new nanoscaled materials(such as apatite or Lamox-type oxides) toimprove their electrical properties at lowtemperature and, on the other hand, tomaterial processing techniques, especiallyin order to reduce the electrolyte thickness(~10 µm) of SOFC.

Plasma spraying is a promising candi-date for such a process because the use ofthis technology has the advantage of

combining the synthesis and consolida-tion of materials. Thermal plasma-basedprocesses indeed offer high reactivemedia to treat or synthesize nanomateri-als. The combination of the plasma char-acteristics and those of precursors(liquid, such as suspensions with dis-persed submicron particles, and/or solid)which are injected within the plasma, canlead to the desired material propertiesand microstructure. For example, denseYttria Stabilized Zirconia (YSZ) elec-trolytes (15-20 µm thick) have been ob-

tained by using Suspension PlasmaSpraying (SPS) (see Figure). A combina-tion of SPS and Solution PrecursorPlasma Spraying (SPPS) allows the dep-osition of finely structured Nickel-YSZcoatings for SOFCs electrode. �

P. Fauchais, R. Etchart-Salas, C. Delbos,M. Tognonvi, V. Rat, J.F .Coudert, and T.Chartier, “Suspension and solutionplasma spraying of finely structuredlayers: Potential application to SOFCs”, J.Phys. D: Appl. Phys. 40, 2394 (2007)

Plasma spraying for solid oxide fuel cells

Intense sources of ultracold particleshave proven to be a very useful tool formetrology and matter wave optics. In thiscontext, the realization of a slow beam ofneutrons through specular reflection onthe blades of a rotor [1] has representeda breakthrough leading to many achieve-ments in neutron optics and interferom-etry experiments. This scheme was alsosuccessfully applied to the slowing downof a supersonic beam of Helium, with the

aim of realizing a very intense source forneutral atom optics experiments [2].

G. Reinaudi and coworkers have re-cently demonstrated the implementationof a similar technique with ultra-cold ru-bidium atoms. In their experimental setup,a laser-cooled atomic packet is sent into a4.5 m long magnetic guide at a velocity of1.7 m/s, and is subsequently slowed downby undergoing an elastic collision with amoving magnetic barrier. This magnetic“mirror” is provided by rare-earth perma-nent magnets mounted on a U-shapedsupport, and set in motion at an adjustablevelocity by means of the mechanical con-veyor belt on which it is fixed (see Figure).

This device permits the removal of upto 95 % of the longitudinal kinetic energy

of the incident atomic packet. This tech-nique was also extended to a set of pack-ets periodically injected into the guide,resulting in the formation of a continu-ous, intense and very slow beam of ultra-cold atoms. Such a beam is a promisingtool for atom-optics experiments includ-ing matter-wave interferometry, and isalso a prerequisite for the achievement ofa continuous “atom laser” [3]. �

G. Reinaudi, Z. Wang, A. Couvert, T.Lahaye, and D. Guéry-Odelin, “A movingmagnetic mirror to slow down a bunch ofatoms’, Eur. Phys. J. D 40, 405 (2006)

References

[1] A. Steyerl, et al.,Phys. Lett. A 116, (1986) 347

[2] E. Narevicius, et al.,Phys. Rev. Lett. 98(2007) 103201

[3] T. Lahaye, et al.,Phys. Rev. Lett. 93 (2004)093003

A Mirror to generate a beam

� Suspension and Solution Precursor Plasma Spraying (respectively SPS and SPPS) – Yttria Stabilized Zirconia(YSZ) coating obtained by SPS and Nickel-YSZ coating obtained the use of a combination of SPS and SPPS

� Figure: Laser-cooled atomic packets are injectedinto a magnetic guide and slowed down by reflec-tion on a moving magnetic barrier. The overlap ofthe successive slowed packets generates an ultra-slow guided beam.

Condensed matter systems, ranging from simple fluids andsolids to complex multi-component materials and even bio-

logical matter, are governed by well understood laws of physics:on the relevant scales of length and time; the appropriatedescription would be just the Schrödinger equation for the quan-tum-many-body problem of the nuclei and electrons interactingwith Coulomb forces. Statistical mechanics would then providethe framework to extend this quantum theory of condensedmatter towards a statistical description in terms of averages takenat nonzero temperature.

However, this program cannot be carried out straightfor-wardly: even dealing only with the associated electrons is stilla challenge for the methods of quantum chemistry. Similarly,standard statistical mechanics makes precise explicit state-ments only on the properties of systems for which the many-body problem can be effectively reduced to a problem ofindependent particles or quasi-particles. Such problems are,for instance, ideal gases, paramagnets, or the multidimensionalharmonic oscillator describing phonons in harmonic crystals.While all these problems are useful and educative (we henceteach them all to students to illustrate the spirit of the generaltheoretical framework) they do not encompass most of the

problems of interest in condensed matter physics: the interac-tions among the considered degrees of freedom introducenontrivial correlations between them. Systematic perturba-tion-theory type methods usually do not lead very far, and un-controlled closed form approximations often fail utterly. Awell-known example is the study of the liquid gas transitionin fluids: simple theories such as the van der Waals equationand its extensions (e.g. the "perturbed chain" statistical asso-ciating fluid theory [PC-SAFT] intended to model the equa-tion of state of polymer melts [1]) give rise to spurious loopsin the isotherms, fail to describe the critical behavior correctly,and may even predict completely unphysical phase equilibriathat do not correspond to any physical behavior of the systembut are mere artefacts of the inappropriate approximations [1].The theory of condensed matter systems contains a never-end-ing list of such failures!

Hence, until about 50 years ago, condensed matter theorysuffered from the basic problem that only a formal frameworkfor the theoretical description in terms of quantum theory andstatistical mechanics did exist, while a reliable set of tools thatwould allow accurate explicit predictions to be made for thestatic and dynamic properties of these systems from first prin-

ciples was simply missing!This unsatisfactory sit-

uation has changed fun-damentally through theinvention of computersimulation, which pro-vides a much more prom-ising approach to theseproblems. Computer sim-ulation is a novel me-thodic route allowing the

Condensed matter theory by computer simulation:from materials to chemical biology [DOI: 10.1051/EPN:2007009]

Mauro Ferrario 1, Giovanni Ciccotti 2 and Kurt Binder 3,1 Dipartimento di Fisica, Università di Modena e Reggio Emilia • Via Campi, 213/A, 41100 Modena • Italy2 Dipartimento di Fisica, INFN, Università di Roma La Sapienza • Piazzale Aldo Moro 2, 00185 Roma • Italy3 Institut für Physik, Johannes Gutenberg-Universitaet • D-55099 Mainz • Germany

� Fig. 1: Typical transition stateconfiguration of the 16 residueterminal fragment of protein G-B1, that forms a so-calledβ-hairpin structure. The twostrands of the backbone, whichis represented as a ribbon, areseparated by a strip of watermolecules. In particular theimportant backbone hydrogenbonds 3-4 (indicated by largespheres) are bridged by watermolecules. Reprinted with per-mission from [2].

europhysicsnews18 • volume 38 • number 3

features

treatment of strong correlations in many-body systems. Start-ing with the introduction of Monte Carlo (MC) methods morethan 50 years ago, and Molecular Dynamics (MD) methodsshortly thereafter, the scope of simulation methods has beenrapidly expanding, and many exciting technical extensions arestill under development, providing broad applications for thisnew theoretical paradigm of science, and new opportunities forimportant discoveries.While in the early days of computer sim-ulation suitable computing facilities were very rare and, com-pared to present-day facilities, very slow and with very smallstorage capacities, very important discoveries were immedi-ately made, which were very surprising at that time. These sur-prises include the fact that hard spheres crystallize at a densitylong before close packing has been achieved, and that dynamiccorrelations in fluids exhibit long time tails. These discoveriesnot only have greatly influenced the thinking of theorists andexperimentalists alike, but have also been the starting point ofa great variety of methodological developments. Now, when apowerful desktop computer is at the hand of every scientist,these advances in simulation methods progress at a breathtak-ing speed! These methods range from basic many-body quan-tum mechanics to classical and quantum-mechanical MC andMD methods, both in equilibrium and far from it, and to tech-niques on a coarse-grained mesoscopic level such as the so-called "Lattice Boltzmann" and "dissipative particle dynamics”methods. The fields to which these methods can be applied areextremely diverse: from fundamental problems in astrophysicsand high energy particle physics (such as the transition fromthe quark gluon plasma to hadronic matter) to basic phenom-ena of solid state physics (such as magnetism, superconductiv-ity and phase transitions between different crystal structures)and to the rich world of phenomena that complex fluids (liquidcrystals, polymers, colloids, micro-emulsions, membranes, etc.)exhibit. Computer simulations in condensed matter now canelucidate properties of materials as well as the major nonequi-librium processes that take place in the living cell.

The school, held at the Ettore Majorana Foundation andCenter for Scientific Culture (EMFCSC), Erice (Sicily), in July2005 emphasized precisely these topics, and was properly en-titled "Computer Simulations in Condensed Matter: FromMa-terials to Chemical Biology". The precise selection of topicsreceived special motivation by the research interests of Prof.Mike Klein (Univ. of Pennsylvania), who was honored at thisevent on the occasion of his 65th birthday. Although in the re-mainder of this article we cannot describe comprehensivelythe contents of the lecture notes of this school (which takeabout 1300 printed pages [2]), we hope nevertheless to sum-

marize concisely where is the front line of current research,using the chapters of these Lecture Notes [2] as a useful guideon this battleground.

Basic tools: classical and quantum simulationsmethodsAs soon as the Metropolis importance sampling method(MC) and molecular dynamics (MD), first proposed in 1953[3] and in 1957 [4], were introduced, they were applied to elu-cidate the equation of state of hard disks. Subsequently thestudy of phase transitions (both of lattice models and of off-lattice systems) has been a very important task of MC andMD methods, since analytical methods are rarely reliable forsuch problems [5-7]. However, a naive application of samplingtechniques to such phenomena will also encounter someproblems: at first order transitions, the states corresponding tothe two competing phases are normally separated by huge freeenergy barriers in phase space [5-7]. Since the sampling al-gorithms do not yield any information on the partition func-tion itself, the information on absolute free energies is lackingand an accurate sampling of the free energy difference be-tween the states is mandatory. This requires (with any stan-dard algorithm) the observation of a large number oftransitions back and forth between the two states. Conversely,at second order transitions the growth of the order parame-ter correlation length (which ultimately diverges when oneapproaches criticality) and the associated slowness of relax-ation ("critical slowing down") have been longstanding ob-stacles to the quest for highly accurate results [5,6].

It has now clearly emerged [2] that there has been impres-sive progress with all these problems. A powerful strategy todeal with the problem of critical slowing down is the applica-tion of the so-called "cluster algorithms": rather than per-forming local Monte Carlo updates (of single degrees offreedom) in the systems, one constructs suitable non-localmoves. These techniques were first introduced for the Isingmodel, where, by the Swendsen-Wang and Wolff cluster algo-rithms (named after the scientists who invented them), oneconstructs large connected clusters of spins having the samesign with a clever recipe. This recipe is derived such that theresulting clusters are independent of their environment andhence can be overturned at zero energy cost. These conceptshave now, to some extent, been carried over to off-lattice sys-tems in the continuum. In this way generic examples such asmixtures of hard spheres with very different sizes, a problemthat hitherto was almost intractable, has become accessible tosolution [2].

features

number 3 • volume 38 • 19europhysicsnews

Also the problem of accurately locating first order phasetransitions and establishing full phase diagrams of compli-cated model systems is now well tractable by Monte Carlomethods. Substantial progress has been achieved by "extendedsampling" strategies. Many of these advanced techniques thatare available today have their roots in the classical "umbrellasampling" [7,8]: the now widely applied "multicanonicalMonte Carlo", "extended ensemble averaging", "multihis-togram methods", etc. Some variations of these methods, suchas "parallel tempering" where a number of systems are run inparallel at a set of neighboring temperatures (or other exter-nal control parameters), which are exchanged from time totime, can also be used in conjunction with Molecular Dy-namics methods (the latter will be discussed below).While insome of these methods reweighting factors need to be esti-mated and this is not always straightforward, a particularlyconvenient approach is the so-called "Wang-Landau sam-pling" [9]: by a rather simple iteration procedure the energydensity of states can be estimated, from which the free energyover the entire temperature range can then be computed. Allthese methods (and many related ones) are concisely de-scribed and put in perspective in Ref. [2]. Applications thatare described range from the estimation of interface free en-ergies to the energy barriers that one encounters in the fold-ing of small proteins.

There is a number of technical problems which are moretypical of classical molecular dynamics [7] than of MC, al-though often they are found possibly in a different shape alsoin MC. The first of these is how to generate averages in en-sembles of statistical mechanics other than the microcanonicalensemble: in other words how to couple the system to thermo-stats, barostats, etc [10,11]. Another typical problem is the si-multaneous occurrence of slow and fast degrees of freedomthat creates difficulties for an efficient use of MD since straight-forward implementations mean a great waste of computer re-sources. Special techniques, such as the freezing of the fastdegrees of freedom by holonomic constraints [12] or multipletime step methods [13], have been developed, starting from theearly times. In Ref. [2] a broad range of issues on MolecularDynamics is spanned, from basic foundations to modern issuesthat are still under investigation, such as the sampling of con-formational equilibria of proteins and free energy barriers.

While the topics discussed so far assume throughout thatclassical statistical mechanics is appropriate, it is clear thatmany physical phenomena of interest require a treatmentbased on quantum mechanics. A special case is the time evo-lution of a quantum particle in a classical environment. Aconsistent description can be based on the Wigner formula-tion of quantum statistical mechanics, allowing the study oftransport phenomena in such mixed quantum-classical sys-tems. Complementary approaches on related problems canbe based on path integral formulations.

At this point, it should be stressed that the path integraltreatment of multi-fermion problems is still hampered by afundamental problem, the "minus sign problem". Some im-provement concerning this problem is probably possible by a

resummation over paths, so the stochastic problem of sam-pling paths is reduced to sampling "graphs". Investigationsalong such lines already present in ref.[2] could well becomea mainstream of research in the future.

Another new type of Quantum Monte Carlo method, the"coupled Ion-Electron Monte Carlo Method", extends thestandard variational Monte Carlo approach [14] to obtain adescription of electronic structure that provides an interest-ing alternative to the density functional based Car Parrinello[15] (CPMD) method. The advantage is that the ions can eas-ily be treated fully quantum-mechanically, and this advantagepays off when one studies problems such as the properties of(metallic) hydrogen under high pressure.

Of course, the density functional theory and CPMD arealso steadily developing further. Ref. [2] provides thoroughreviews of these techniques, from their basic principles to thecurrent state of the art. Particularly interesting extensionsconcentrate on the possibility to accelerate rare events andcompute free energy barriers.

The sampling of rare events, e.g. a nucleation processwhere a huge free energy barrier needs to be crossed to forma critical nucleus of the new stable phase on the backgroundof a metastable phase is a longstanding challenge to simula-tion. Transition path sampling has been the first scheme in-troduced to compute the relevant properties withoutassuming a reaction coordinate. In Ref. [2], both the formaltheoretical background ("transition path theory") of this ap-proach and the state of the art of the implementation of thisnew techniques are discussed, as well as applications to studythe pathway of protein folding. (Fig. 1)

Selected applications,from simple fluids to human immune responseThis article can only attempt to convey to the reader theflavour of the wide range of problems that one can addresstoday via computer simulation in the science of condensedmatter: while in many cases a reductionist approach is needed,progress is achieved, in terms of coarse-grained models, bythe imaginative ideas of scientists suitably adapting new mod-els to the problems under study. Coarse-graining means thatsome small-scale degrees of freedom, which are consideredless relevant to the questions considered, are eliminated orelse treated in simplified ways. Computer simulation offersthe advantage that connections can be established betweenthe models of condensed matter on different scales and thehierarchy from the sub-Angstrom scale, where one deals witheffects due to the electrons, up to the mesoscopic and macro-scopic scales relevant for living matter.

There are cases where a well-defined separation of scalesoccurs between relevant and irrelevant degrees of freedom(e.g. colloidal suspensions) and others where this is not thecase (e.g. polymer melts, or biological membranes). In col-loidal suspensions, particles of spherical or rod-like or disk-like shape with linear dimensions in the micrometer range,are immersed in a solvent fluid, which acts as a heat bath forthe larger colloid particles (but presents also hydrodynamic


features

interactions, and, via ions or polymers dissolved in the fluid,controls also the effective interactions between the colloids).This is a classical case [2] where all the degrees of freedom ofthe system apart from those of the colloids can be integratedout, and models result which resemble models of simple flu-ids (the dynamics of the colloidal particles being, however,rather different).

The situation is very different for polymeric systems: a flex-ible macromolecule exhibits nontrivial structure from theAngstrom scale (chemical bonds) over the coil size (tens ofnanometers) to even larger length scales of mesoscale orderingphenomena. Consequently, there does not exist a unique mod-elling approach to deal with matter formed from such macro-molecules. Rather there exists a large diversity of proposedalgorithms and different models, as outlined in depth in Ref. [2],to confront fundamental problems of polymer science, such as,for example, understanding which material properties controlthe "tube diameter" describing the constrained diffusion of poly-mers in melt (note that polymers "creep" in a snakelike motionalong a "tube" that follows their own contour). (Fig 2)

Of course, there are still limitations that have not yet beenovercome: while the simulations havecontributed much towards the under-standing of the slowing down of under-cooled simple fluids near the glasstransition and while one can also studyglassy freezing of very short, non-entan-gled polymers, the interplay of chain en-tanglement and glassy relaxation clearlyis beyond reach. Similar caveats apply tothe studies of spreading and bridging ofpolymer droplets on substrates - there isstill a huge gap between the nanometerlength scale and nanosecond time scale ofthe simulations on the one hand, and themacroscopic scales of the experiment onthe other hand. Similar caveats apply tomany other problems also discussed in [2],such as defect patterns in liquid crystallinesystems, surfactant layers under shear, lipidbilayers with transmembrane proteinsmodelling biological membranes, etc.However, one should emphasize that forproblems of this type, apart from the simu-lations, there is no reliable theoretical in-formation whatsoever. (Fig 3)

Even in the "classical" problem of spinglasses, a problem heavily debated and wellstudied for more than thirty years, there isstill no relevant analytic theory apart from the mean-field theory of the infinite-range model where every spin inter-acts with every other spin with random Gaussian couplings.

Experiments on these systems can neither probe the orderparameter distribution nor the relevant correlations directly,so the guidance from simulations has been absolutely crucialfor a better understanding of this problem.

Particularly challenging are the simulation tasks in bio-logically motivated physics: Drug-target binding necessitatesquantum-mechanical approaches for the chemically reactivemolecular groups, but the huge size of complex biomoleculesin aqueous solution requires a simplified description on longlength scales. So hybrid approaches have been introduced thatcombine ab initio with classical MD in the environment ofthe reactive groups, by suitable "quantum mechanical/mole-cular mechanical (QM/MM)" partitioning.

While in this problem one tries to deal with atomistic de-tail, the problem of evolutionary design of the so-called T-cell receptors that recognize and eliminate tumor cellsrequires a completely abstract model with no chemical andstructural details [2].

OutlookWe have already emphasized the need to develop mixed (abinitio/classical MD) approaches, where a subsystem is treatedab-initio, and coupled to a classical environment via a transitionregion where a proper "handshaking" between •••


features

� Fig. 2:Result of the “primitive path”analysis for allchains in the melt of 100 bisphenole-A polycarbonate chains of N=60repeat units. The primitive path highlights the geometry of chain entangle-ment. The red chain is entangled with all those chains whose primitive pathsare shown in blue, but not with those shown in light gray. Reprinted with per-mission from Macromolecules 38, p. 8089(205).

••• both types of simulation is attempted. This is an example of"multiscale simulation", an important paradigm of the field: tocombine, in a coherent fashion, complementary methods ofcomputer simulation on vastly different time and lengthscales for problems of capital interest. e.g., for multi-func-tional materials hierarchically organized from complex build-ing blocks a molecular simulation may be impossible becauseof the excessive need for computer time. When one then re-sorts to coarse-grained models, however, there is a clear needto validate the coarse-grained model by a "mapping" to anatomistic model on smaller scales. In fact, there may also bethe need to reintroduce atomistic detail consistently intocoarse-grained models ("inverse mapping"). Such techniqueshave been proposed in the context of polymer melts, but it isclear that the problem is much more general, and far fromfully solved. The goal is to develop simulation approaches forcomplex materials and biological matter that successfullybridge the gap from the small scales of electronic structurecalculations to the mesoscopic scales of pattern formation insoft matter (where one uses coarse-grained simulation tech-niques such as dissipative particle dynamics, multiscale colli-

sion dynamics, etc.). While this goal will remain an excitingchallenge for many years to come, nevertheless steady and im-portant progress towards reaching this goal can be antici-pated. Thus, we can expect that the role of computersimulation as the central and basic method of approach ofcondensed matter theory will become even much more im-portant in the future. �

About the authors:Mauro Ferrario is Professor of Structure of Matter at the Uni-versity of Modena. He is a computational physicist workingon interdisciplinary problems between physics and chemistry.Giovanni Ciccotti is Professor of Structure of Matter at theUniversity of Roma "La Sapienza". He has contributed to thedevelopment of Molecular Dynamics methods now for sev-eral decades.Kurt Binder is Professor of Theoretical Physics at the JohannesGutenberg Universitaat of Mainz. He belongs to the "oldhorses" generation in the field of Monte Carlo simulations andwill receive this year the Boltzmann Medal in the field.

References[1] L.Yelash, M. Müller,W. Paul, and K. Binder,PCCP 7, 3728-3733 (2005)

[2] M Ferrario, G. Ciccotti and K. Binder (eds.) Computer Simulationsin Condensed Matter: From Materials to Chemical Biology, Vols 1,2Springer, Berlin-Heidelberg (2006)

[3] N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, andE. Teller,Chem. Phys. 21, 1087 (1953)

[4] B. J. Alder and T. E. Wainwrigt, in Proceedings of the InternationalSymposium on Statistical Mechanical Theory of Transport Processes,Brussels 1956, ed. I. Prigogine, Interscience, NY (1958); see also, fora collection of the reprints of the papers which made the field, thebook: G. Ciccotti, D. Frenkel and I. R. MacDonald (eds.) Simula-tion of Liquids and Solids. MD and MC Methods in StatisticalMechanics, North Holland, Amsterdam (1986)

[5] K. Binder "Monte Carlo Investigations of Phase Transitions and Criti-cal Phenomena" in C. Domb and M. S. Green (eds.) PhaseTransitions and Critical Phenomena, Vol 5b, pp. 1-105. AcademicPress, New York (1976)

[6] D. P. Landau and K. Binder "A Guide to Monte Carlo Simulations inStatistical Physics", 2nd ed. Cambridge, Univ. Press, Cambridge (2005)

[7] G. Ciccotti and W. G. Hoover (eds) "Molecular Dynamics Simulation ofStatistical Mechanical Systems", North Holland,Amsterdam (1986)

[8] G. M. Torrie and J. P. Valleau, J. Comp. Phys. 23, 187 (1977)

[9] F. Wang and D. P. Landau, Phys. Rev. E 64, 056101 (2001)

[10] H. C. Andersen, J. Chem. Phys. 72, 2384 (1980)

[11] S. Nosé,Mol. Phys. 52, 255 (1984)

[12] J. P. Ryckaert, G. Ciccotti and H. J. C. Berendsen, J. Comp. Phys. 23,327 (1977)

[13] M. Tuckermann, B. J. Berne and G. J. Martyna, J. Chem Phys. 97,1990 (1992)

[14] B. L. Hammond, W. A. Lester, Jr. and P. J. Reynolds "Monte Carlomethods in Ab Initio Quantum Chemistry"World Scientific,Singapore (1994)

[15] R. Car and M. Parrinello, Phys. Rev. Lett. 55, 2471 (1985)


� Fig. 3: Typical peptide configurations embedded in a coarse-grainedmodel of a biological membrane, for three different peptide sizes. Reprintedwith permission from [2].

features

One of the present au-thors essayed a lecture [6]on such teasing trivialitiesat the “International Con-ference on Theoreticaland Applied Mechanics” inWarsaw (2004). An eageraudience overflowed thecapacity of the lecture the-atre. At the preceding con-gress in Chicago (2000), aspart of a Science TeachersDay, Professor RaymondC. Turner of Clemson Uni-versity told an appreciativecrowd of local physics andscience teachers aboutusing toys to teach. Manyin the audience were veryconversant with this modeof instruction, but keen tolearn more.

Childlike absorptionMany famous personalities of physics have had a taste for toysand tricks. Von Kármán [7] described Prandtl's “childlike ab-sorption”:In subsequent months I came to know Prandtl quite well. Hewas a man of strange contrasts. Sophisticated and gifted in sci-ence, and a great teacher on a person-to-person basis, he wasnonetheless naïve about life and childlike in behavior.He couldnot pass a toy in a shop, for instance, without curiously finger-ing it, and he was spell-bound by routine magicians tricks.Once at a party the guests decided out of fun to test Prandtl'swell-known love of toys. In one corner of the room somebodyplaced a child's gyro, in such a position that it would be in fullview. Everybody awaited Prandtl's arrival to see whether theprofessor would live up to expectations and go directly to thetoy.He didn't disappoint them.As soon as he entered the roomand his eye fell on the toy, he had no interest in anything elseor in anyone at the party. It was almost half an hour beforePrandtl realized that the party had formed around him andthat everyone was secretly enjoying his childlike absorption.Other well-known scientists who have written about the analy-

sis of toys and games include G.-G. Coriolis, today principallyknown for the acceleration that bears his name, but a distin-guished mechanician who in 1835 wrote“Théorie mathématiquedes effets du jeu de billard”; F. Klein and A. Sommerfeld, •••

In the early 19th century the popularisation of science becamea vogue, expressed in books that often took the form of “con-

versations on natural philosophy”. They may have echoed thediscourse of the classical dialogue of debate but this was morethe conversation of the drawing-room, albeit in the stilted parl-ance of the time.The top, my boy is a subject which the great Mantuan barddid not consider beneath the patronage of his muse. (anon.,quoted in [1])Such lessons were aimed at future scientists (“boys”), the

workers (“operatives”) and even the gentler sex (“little house-wives”). Of three particularly influential texts [1,2,3], two wereby women [2,3], one of whom preferred to remain anonymousfor some time. The third, written by the early cancer researcherJohn Ayrton Paris, was splendidly entitled“Philosophy in Sportmade Science in Action, being an Attempt to implant in theYoung Mind the first Principles of Natural Philosophy by theaid of Popular Toys and Sports of Youth”. Examples of the manyplaythings included as “instruments of philosophical instruc-tion” are the seesaw, the kite, the top, and the Jew's Harp.

“And is it then possible'', said the vicar, in a tone of supplica-tion, “that you can seriously entertain such a wild and, I mightadd, kill-joy scheme?” [1]Paris' kill-joy scheme ran to many editions and imitations and

helped found a pleasant tradition that was followed in later yearsby such popularisers as C.V. Boys in his lectures on soap bubbles,an early version carrying the flowery title “Soap-bubbles, theirColours and the Forces which mould them: Being the Substanceof many Lectures delivered to Juvenile and Popular Audienceswith the Addition of several new and Original Sections” [4]. Per-haps it faltered in the period of high seriousness about physics inthe mid-20th century; nuclear physics had no place in the nurs-ery. But today it emerges afresh in the minds of our physicsteachers at every level (for a recent collection of experiments onsoap bubbles see the books by Rämme [5]). Museum shops offera variety of such diversions, from cheap plastic to executive-toychrome, but many can be found in the kitchen, the toolbox and

the sewing cabinet, or are easily con-structed. Toys have become part of ourcultural heritage, as one observes in ar-chaeological digs, in the items for salein antique shops, in museums, and theinclusion of the “slinky” in the USPostal Services “Celebrate the Cen-tury” series of postage stamps.


features

Toying with physics [DOI: 10.1051/EPN:2007010]

Hassan Aref 1, Stefan Hutzler 2 and Denis Weaire 2,1 Department of Physics, Technical University of Denmark • 2800 Kgs. Lyngby • Denmark (Permanent address: Department ofEngineering Science and Mechanics, Virginia Tech, Blacksburg • VA 24061 • USA)2 School of Physics, Trinity College Dublin • Dublin 2 • Ireland

� Fig. 2: A page from Jellett's Treatiseon the Theory of Friction (Macmillan,London, 1872) where Jellett's constantis introduced.

� Fig. 1: Slinky on a US postage stampcommemorating the ``Slinky craze'' of 1945.

europhysicsnews

••• seminal figures of the late 19th and early 20th century in math-ematics and physics, respectively, who collaborated on “Überdie Theorie des Kreisels”, publishing all four volumes in Leipzigbetween 1897 and 1910; H. Bondi, J. L. Synge, and K. Stewart-son all wrote about spinning tops and so on. This tradition ofstudying toys seriously has, fortunately, survived to the presentday, and we find in the current literature and in premier jour-nals contributions on the “Levitron” [8], on “Newton's cradle”[9] (and references therein), and on “Euler's disk” [10].

MaddeningmechanicsWith so much to choose from, we shall concentrate on me-chanics, a subject misleadingly presented as rather trivial in el-ementary physics classes.

The tractability of many problems of rational mechanicsmay be its attraction: even the spinning rigid body bows to the-ory for the special cases usually considered. But add a bit ofasymmetry, or a dash of friction or other contact forces, andyou soon have a recipe for strange, counter-intuitive gravity-defying behavior. Unpalatable mathematical formalism isneeded to explain it.

Many effects were well known already in the nineteenth cen-tury. For example, a spinning top will rise to the vertical “sleep-ing” position: why? In his popular work “Spinning Tops” JohnPerry (1890) waxed sarcastic on the subject of Cambridge menwho thought they surely understood but had forgotten the de-tails. It had something to do with the action of friction on therounded tip of the top - it is fairly evident that reducing it to apoint removes the hope of an explanation. For a proper eluci-dation Perry was directed by another Irish phyicist, GeorgeFrancis Fitzgerald, to the work of his father-in-law. John Jellett's“Treatise on the Theory of Friction” contains much wisdom onthe role of friction that endures today, in particular “Jellett'sconstant”.

The constant arrives unheralded in a section of miscella-neous applications (see Fig. 2). It is the result of judicious ap-proximation: if the rotation about the symmetry axis of the top

is regarded as dominant, an additional conserved quantityemerges. Jellett used it in a neat argument to show that, as fric-tion reduces the energy of the top, it must rapidly bring it tothe upright position.

But with both friction and asymmetry we are still in real trou-ble. The classic example is the celt (pronounced“selt”), rattlebackor wobblestone, guaranteed to stun a class that sees it reverse itsdirection of rotation on an overhead projector. In his inspiringbook “The Flying Circus of Physics” Jearl Walker writes [11]:Some of the stone instruments made by primitive men... dis-play curious personalities...These stones, called celts, are gen-erally ellipsoidal in shape. When you spin them about avertical axis some behave as you would guess, but others actnormally only when spun in one direction about the vertical.If you spin them in the other sense, the rebellious stones willslow to a stop, rock for a few seconds, and then spin in theirpreferred direction...If you tap one of these stones on an end,... it will rock for a while. But soon the rocking ceases, and thestone begins to rotate about the vertical axis.What causes suchpersonalities?Figure 3 depicts a neat variation.A“tunable” rattleback, com-

mercially known as the Russian rattleback (probably becauseof the style of decoration). The two small turtles may be rotatedon the “deck” of the otherwise completely symmetric “hull”.When they face one another, the object rotates with equal easein the clockwise or the counterclockwise direction. However,turn the turtles so that they are perpendicular to the axis of thehull, facing in opposite directions, and the mysterious behavioremerges at once. They like to ride in the forward direction,and wobble and reverse direction if spun so they are ridingbackwards.

Perry describes another variation: an ellipsoidal stone spunon a surface, about one of its shorter axes: it staggers into theupright position, spinning about its long axis. A further elabo-ration is the additional end-to-end asymmetry of an egg, which,if spun with its blunt end down will flip over, and spin on itssharp end. This is the same behavior seen in the “Tippe-Top”toy, which has fascinated amateur and professional physicistalike. A famous snapshot (Fig. 4) shows Wolfgang Pauli andNiels Bohr experimenting with the apparent angular-momen-tum-reversing behavior of the tippe-top.

According to Perry, Lord Kelvin would experiment withoddly shaped stones when he went to the seaside, exploringsuch peculiar dynamics. He was also fond of eggs, for under-graduate demonstrations, baffling the class with the differentproperties of identical eggs (one of which was hard-boiled).

In a modern study of eggs worthy in its erudition of anemeritus Cambridge professor, Keith Moffatt has addressed notonly some of these extraordinary feats, but also the astonishingcapacity of an egg to jump vertically in the course of its rota-tional manoeuvres [12].

Moffatt came to the problem via a simpler one, that of theEuler disk. This is just a metal disk, the heavier the better.Whenspun on a flat surface it gradually slows, but its audible fre-quency of rotation increases dramatically. It rises inexorably to-wards a crescendo and stops abruptly, like the best symphonies.


� Fig. 3: The so-called Russian rattleback in which the two small controlmasses (shaped as turtles) can be adjusted to create an asymmetry that dra-matically affects it motion.

features

Who said dissipation causes motion to die away slowly? Hereit provokes a mathematical crisis called a finite-time singular-ity, important in modern fluid dynamics.

In relatively modern times, the rotational toy that has madethe biggest move in the stock market is surely the “yo-yo”. Likemany of our toys it has a pedigree measured in millennia, andthere have been periodic crazes in the modern era. (Investnow.) Another device of ancient origin is the diabolo. JamesClerk Maxwell was given one as a child and took it with him toCambridge (Fig. 5).

All of the above is certainly abstruse - some would say that itis more of a lesson in humility for the teacher than elementaryphysics for the pupil. So let us retreat to something simpler -“Newton's cradle”, as shown in Figure 6. A ten-second experi-ment with a ten-second explanation in terms of conservation ofenergy and angular momentum? But there is a greater lesson tobe learned: do the experiment, and make careful observations...

We observe that, if one ball is pulled aside and released, it isnot precisely true that only one ball is ejected from the otherside. Merely some misalignment in manufacture? It turns outthat the cause is not to be found there, but rather in the finitecompressibility of the balls, which renders collisions non-in-stantaneous. A compressive elastic pulse of finite width is prop-agated down the line. When it ejects the last ball, the previoustwo are still slightly compressed together, so they fly apart tooin a clearly visible displacement (the remaining displacementsare not usually discernible). Recognizing this, we see that thesubsequent motion cannot long conform to what is expected(and taught by many). Indeed it does not, and several papersaddress this motion; see [9] and referencestherein. To observe it unmodified by airdrag, build a big Cradle, and rock it. Dis-carded bowling balls would be good.

Funny fluidsFluid mechanics has its share of toys, exec-utive and otherwise. The “Ooze Tube” andits variants allow one to watch the coiling ofa thread of viscous liquid falling through asmall circular hole, see Figure 7. Barnes andWoodcock [13] call this the “liquid rope-coil effect”. In his paper at the “12th Interna-tional Congress of Applied Mechanics”, heldat Stanford University in 1968, the father ofmodern fluid dynamics, G. I. Taylor showedand analyzed examples of instabilities ofjets, threads and sheets of viscous fluids.

The thin thread with a circular cross-section coils neatly into a growing mound.

A flat ribbon, on the other hand, will fall in layers. Whenplaced horizontally, and thus constrained to move two-dimensionally, the thread will buckle. Taylor gives the follow-ing explanation [14]:The reason for the instability is clear. If the stream is very thin,the longitudinal compression acts in the same way as end com-pression in a thin elastic rod. The rod becomes unstable at acertain load, and less force is needed to move the ends towardsone another when the rod is bent than when it is straight. Thisis called Euler instability.Experiments [13] show that the effective height of fall and

the frequency of coiling of a viscous thread are proportional.Is this clear from elementary considerations? Simple dimen-sional analysis, using just the coiling frequency, the height offall, and the acceleration of gravity, inevitably yield a frequencyinversely proportional to the square root of the height - not atall the observed relationship. However, introduce the kine-matic viscosity of the liquid, as would immediately be sug-gested by simple fluid mechanics, and dimensional analysisyields two dimensional groups, one of which is a “Reynoldsnumber” for the flow. For small values of this dimensionlessnumber or slow flow, we obtain a coiling frequency propor-tional to the fall height as observed.

The so-called “Brazil nut effect” is the observation that in amixture of granular material of different sizes, e.g., a can ofmixed nuts, the large particles are often observed to end up ontop. One intuitive explanation is that when the container isshaken, the small particles flow to fill in voids with greaterprobability and facility than large particles, hence achieving •••


features

� Fig. 4: Two giants of modern physics, WolfgangPauli and Niels Bohr, fascinated by the counter-intu-itive behavior of the tippe-top. (Photograph by ErikGustafson, courtesy AIP Emilio Segre Visual Archives,Margrethe Bohr Collection) (reproduced with kindpermission of AIP).

� Fig. 7: Fluid mechanical toys. up, the“Ooze Tube” illustrating the instability of aviscous liquid thread. Center, the “Sandwand”- how does one move the steel ballto the to top of the granular material inthe tube? down, the “Cartesian diver” illus-trating simple principles of buoyancy.

••• segregation. There is a me-chanical toy, the “Sand wand”,that exploits this effect, seeFigure 7. Enclosed in a trans-parent cylindrical tube is aroughly monodisperse gran-ular material and a singlesteel ball that almost spansthe tube. Move the ball, one isinstructed, to one end of thetube. Gentle manipulation,rotation, tapping, and so onwill not do the trick. A fewsturdy shakes with the rod es-sentially vertical and the de-sired result is achieved, theglittering ball emerging fromthe smaller grains and settlingon top of them.

A perennial favorite is the“Cartesian diver” or some-times Cartesian devil - al-though the attribution toDescartes is, apparently, mis-placed - often artfully craftedas a small demon-like creature,submerged in a vertical glasscylinder filled with water, seeFigure 7. The cylinder has aflexible membrane at the top.

Depress the membrane and the diver descends as if magicallyheeding instructions from a distance. Release the pressure andhe ascends. With a bit of practice he can be made to swirl anddance in the water column. What is the source of this magic?The diver is constructed with a small air-filled chamber inside,open at the bottom and there in contact with the water. Whenthe water column is compressed, the air pocket within the diveris much more easily compressed. The buoyancy of the diver(and the air pocket with him) is reduced and he descends. Aswith the rest of us, already struggling to stay afloat, a bit morepressure applied from above and we are totally submerged!

Positive playThere is a creative playfulness to good science, and using toysin instruction seems to bring this full circle. They are ap-proachable, intuitive, familiar, we are tempted to say “easy”. Wefeel that we understand how they work.We certainly know howwe expect them to work. When they turn out to challenge ourintuition and our deepest analytical abilities, do we become in-secure or intrigued? Hopefully the latter and, hopefully, thatcuriosity breeds interest, engagement, creativity, and, ultimatelyan affection for scientific inquiry. �

About the authorsHassan Aref is the Reynolds Metals Professor of EngineeringScience and Mechanics at Virginia Tech, and the Niels Bohr

Visiting Professor at the Technical University of Denmark. Hisinterests include fluid mechanics, dynamical systems, appliedmathematics and mechanics education.Denis Weaire is Erasmus Smith Professor of Natural and Ex-perimental Philosophy at Trinity College Dublin and has akeen interest in the history of Physics. Together with StefanHutzler, who is a lecturer at TCD, he has written a book onthe Physics of Foams, containing further examples how child-like fascination (in this case with soap bubbles) can be turnedinto science.

References[1] J. A. Paris,Philosophy in Sport Made Science in Earnest, Longman,

London (1827).

[2] J. Marcet,Conversations in Natural Philosophy, Longman, London(1819).

[3] M. Edgeworth, Harry and Lucy Concluded,Hunter, London (1825).

[4] C. V. Boys, Soap-Bubbles, their colors and forces which mold them(1890) (Dover Science Books, 1958).

[5] G. Rämme, Soap Bubbles in Art and Education, Willie Yong, Singapore(1998).Experiments with Soap Bubbles and Soap Films, Uppsala,Sweden (ISBN 91-631-6999-1) (2006).

[6] H. Aref, “Toys and games in mechanics education.” Paper FSM7L-10003, XXI International Congress on Theoretical and AppliedMechanics, Warsaw, Poland, 15-21August 2004 (2004).

[7] Th. von Kármán and L. Edson, L.1967 TheWind and Beyond, Little& Brown, Boston-Toronto, (1967).

[8] M.V. Berry,Proc. Royal Soc. (Lon-don) A 452,1207 (1996).

[9] S. Hutzler, G. Delaney, D. Weaire andF. MacLeod,Am. J. Phys. 72 1508(2004).

[10] H. K. Moffatt,Nature 404, 833(2000).

[11] J. Walker,The Flying Circus ofPhysics, (2nd edition 2006), JohnWiley and Sons (1975).

[12] H.K. Moffatt, and Y. Shimomura,Nature 416, 385 (2002).

[13] G. Barnes and R. Woodcock,Am. J.Phys. 26, 205 (1958).

[14] G. I. Taylor, “Instability of jets,threads, and sheets of viscous flu-ids.” Proceedings of the 12th

International Congress of AppliedMechanics (Stanford, 1968),Springer, pp.382-388. (1969).


features

� Fig. 5: Maxwell's diabolo. (Pho-tograph courtesy of the Universityof St. Andrews.)

� Fig. 6: Newton's cradle.

In the beginning of the 19th century, T.Young and J. Fraunhoferbuilt the first optical diffraction gratings and showed the role

of optical diffraction in their behaviour. Since that time, diffrac-tion gratings have been used in a broad range of technologicalapplications, from the earlier accurate optical spectrometers tothe recent integrated optical devices. They are also used in abroad range of wavelength from X-ray to microwaves. In thiscontext, diffraction gratings have been the cornerstones ofmany publications for almost two centuries. Surprisingly, nowa-days the subject is far from being exhausted.

Every student knows that a grating can spread a white lightbeam from an incandescent lamp into a continuous spectrum ofcolors.For a grating made with a one-dimensional lattice of thick-thin wires, this‘’rainbow’’ is duplicated many times,each one cor-responding to a diffraction orderm (See Fig.1 for instance).Thesebasic behaviours of diffraction gratings can be easily understoodby the conservation of light momentum such as

k→→

s = k→→

i + K→→

(1)

where k→→

s is the momentum of the scattered light, k→→

i is the mo-mentum of the incident light, and K

→→

is a vector of the recipro-cal lattice of the grating such that K

→→

•a→→ = 2πm(m��), where a→→

is the basis vector of the lattice. The general behaviour of opti-cal grating results from this basic assumption. For instance, inFig.1, one considers a reflecting grating. An incident light i isthen scattered into three diffraction orders. The zero order(m= 0) is known as the specular beam. In Fig.1, two other or-ders are considered, m= 1 and m= -1.

Wood anomaliesIn 1902, R. W. Wood observed some unexpected patterns in thespectrum of light resolved by optical diffraction gratings [1].The Wood spectrum presented many unusual rapid variationsof its intensity in certain narrow wavelengths bands. As a con-sequence, these effects unexplained by ordinary grating theorywere named ‘’Wood anomalies’’.

In 1907, Lord Rayleigh proposed an explanation of thoseanomalies [2]. Considering for instance the case of the Fig.1,one notes that the order m= 1 becomes tangent to the surfacegrating before disappearing. In such a case, the light momen-tum component along the Oz axis of a diffracted order m, i.e.kz,m, becomes imaginary right after having been cancelled.Then, the diffracted order becomes evanescent (non-homoge-neous order). When kz,m is real, the diffracted order is free topropagate along the Oz axis (homogeneous order). In the pres-ent case, as the order m= 1 turns to be non-homogeneous, itsenergy will be redistributed over the other orders. As a conse-

quence, the diffracted beam intensity of the specular beam(m= 0), for instance, increases just as the diffracted order m= 1vanishes. For an incidence θ (see Fig.1), a diffracted order mbecomes non-homogeneous for wavelengths greater than aspecific value, the Rayleigh wavelength, for which an anomalyoccurs. One then talks about a Rayleigh anomaly.

Around 1938, U. Fano proposed another explanation [3],where the anomalies are related to a resonance effect. Such a res-onance comes from a coupling between an eigenmode of thegrating and a non-homogeneous diffraction order (one also talksabout a resonant diffraction order), i.e. after it vanishes for wave-lengths greater than the corresponding Rayleigh wavelength.

From an experimental point of view, gratings can then showsome very complex behaviour often hard to explain in detailswithout the use of Maxwell’s equations. In addition, solvingMaxwell’s equations in order to study the electromagnetic dif-fraction quickly becomes a difficult problem, even with simplegeometries. Except for some simple and powerful analyticalworks, such as those from Lord Rayleigh or U. Fano, it will benecessary to wait until the Sixties and the beginning of com-puter calculation to observe significant progress in the Woodanomalies comprehension. For instance, one can underline thefirst fundamental numerical results of Hessel and Oliner in1965 [4], and Maystre and Nevière in 1977 [5,6]. Hessel andOliner presented a wide study of the eigenmode resonance rolein Wood anomalies [4]. They have shown that depending onthe type of periodic structure, the two kinds of anomalies i.e.,Rayleigh anomalies or resonant anomalies may occur sepa-rately or are almost superimposed.


features

Bounded modes to the rescue of optical transmission [DOI: 10.1051/EPN:2007011]

Michaël Sarrazin and Jean-Pol Vigneron, Laboratoire de Physique du Solide, Département de Physique • FUNDP • 61 rue de Bruxelles • B-5000 Namur • Belgium E-mail : [email protected]

� Fig. 1: A reflecting optical grating. i is the incident light beam. The light beamslabelled m are some of the diffracted orders and θ is the angle of incidence.

Maystre and Nevière studied specific cases of such resonantanomalies [5,6]. For instance, they presented a wide study about’’plasmon anomalies’’, which occurs when the surface plasmonsof a metallic grating are excited [5]. They also considered ananomaly that appears when a dielectric coating is deposited ona metallic grating, and corresponds to guided modes reso-nances in the dielectric layer [6].

It is important to underline that, in some recent publications,some authors improperly reduce the Wood anomalies to theRayleigh anomalies only. It is important to recall that the Woodanomaly concept covers the resonant anomalies (Fano anom-alies) and the Rayleigh anomalies at the same time. In addition,the resonant part of the phenomenon can imply many kinds ofresonances and not only the surface plasmons resonances. Amazingly, thirty years after the pioneer works, in the early 21st

century, the experimental and theoretical interest in these sub-jects has not dried up.

Optical transmission in a nanostructured slab In 1998, Ebbesen et al [7] reported on optical transmission ex-periments performed on periodic arrays of subwavelengthcylindrical holes drilled in a thin metallic layer deposited onglass (see Fig.2). These experiments renewed the motivationfor investigating metallic gratings, in particular those with atwo-dimensional lattice. The most attractive characteristic oftheir results was the peculiar wavelength dependence of thetransmission (see Fig.3). The latter was defined as the ratio be-tween the energy of the specular transmitted beam (m= 0) onlyand the energy of the incident beam. After these first experi-mental observations, the role of the thin metallic film surfaceplasmons (SPs) was put forward in order to explain the pecu-liar wavelength dependence of the transmission [7].

A surface plasmon on a plane surface is a non-radiative elec-tromagnetic mode, associated with a collective excitation of elec-trons at the interface between a conductor and an insulator. Inthis way, a surface plasmon cannot be excited directly by light,and it cannot decay spontaneously into photons. The non-ra-diative nature of SPs is due to the fact that interaction betweenlight and SPs cannot simultaneously satisfy energy and momen-tum conservation laws. Moreover, the momentum conservationrequirement can be achieved by roughening or corrugating

the metal surface, forinstance. It is exactlythe situation in thosedevices. In this case,the transmission peakswere interpreted asSPs resonances [7]. Sur-prisingly, this approachseems to have been sug-gested independently ofthe knowledge of some

earlier works about the reflecting one-dimensional gratings.Then, the first attempt to understand the Ebbesen experimentshas missed some important earlier results of the optical grat-ings study.

In this way, in these first interpretations, the exact role of SPswas not clearly assessed, and many questions remained to clarifycompletely the scattering processes involved in these experiments.

As a consequence, some other explanations had been devel-oped which could compete with the SPs model. For instance, itwas suggested that these phenomena could also be described interms of cavity resonances taking place into the holes [8]. Morerecently, T. Thio and H. Lezec had also suggested a diffractedevanescent wave model [9].

In addition, some doubts against the SPs hypothesis reliedon some experimental results about subwavelength hole arraysmade in non-metallic films. Though SPs cannot exist, the trans-mission pattern was found to be very similar to that obtainedwith a metallic film [9]. Surprisingly, the very fact that the typ-ical transmission pattern can be observed even in non-metal-lic systems convinced some authors to fully reject the SPshypothesis and rather consider models involving nonresonantevanescent waves diffraction. To our knowledge, this point ofview is not supported by recent results. In fact, the SPs must bereplaced by other kinds of eigenmodes in the nonmetallic cases.

As a consequence, despite alternative theoretical interpreta-tions, a large experimental consensus, many theoretical results,tend to prove that the SPs interpretation was correct in themetallic films cases, and must be extended to a bounded modeinterpretation in the most general cases [10-12]. It is only thetransposition of the 1960’s results about one-dimensional re-flection gratings to bidimensional transmission gratings.

Let us clarify this.

Bounded modes and transmissionBasically, the gratings considered here are simply drilled slabs.In this case, a bounded mode, i.e. an eigenmode of the grating,is an electromagnetic field configuration such that the field istrapped along the slab. The mode can propagate along the slab,but not elsewhere. The amplitude of the field decays then, ex-ponentially from each side of the slab. Typically, two amplitudepatterns exist, which correspond to symmetric and antisym-metric modes. As the thickness of the slab increases, in somecases, the modes can refer to pure surface modes (Fig.4), i.e.both slab interfaces are decoupled. Surface plasmons are suchmodes for instance.

As previously suggested, bounded modes cannot be exciteddirectly by incident light or decay spontaneously into photons.Moreover, the coupling between the outside electromagneticfield and the bounded modes is allowed by the use of rough orcorrugated surface, as in the present case. The eigenmodes, andtheir coupling with outside light, can be described mathemat-ically as follow.

As in quantum mechanics, a scattering problem can betreated in electromagnetism via the use of the scattering matrixformalism (S matrix). In such a representation, it is assumedthat the electromagnetic field can be described by two super-


features

� Fig. 2: The typical bidimensional array of subwavelength holes.

Thin optical absorbing slab

Holes

Dielectric substrate (with an imagi-nary part of the permittivity almostequal to zero

vectors IFin>> and IFscatt>> (with the use of ‘‘ket’’ representation),which correspond to the incident electromagnetic field that litthe studied device and to the field scattered by the device. Forinstance, each component of these vectors corresponds to aspecific diffraction order. Then, incident and scattered field arelinked via the scattering matrix defined as

IFscatt>>= S(λ)IFin>> (2)

so that S(λ) contains all the physical information about thestudied device, such as a diffraction grating. In addition, it canbe shown that S(λ) can be explicitly defined from the Maxwell’sequations. For instance, the simulations in our papers are basedon such a method, which combines scattering matrix formal-ism with a plane wave representation of the fields. This tech-nique provides a computation scheme for the amplitude andpolarization (s or p) of reflected and transmitted fields in anydiffracted order.

A fundamental key to the understanding of diffraction grat-ing phenomenology is to define the eigenmodes of the studieddevice. Eigenmodes then obey the homogeneous equation

S-1(λp)IFeigen>>= 0 (3)

in such a way that the eigenmodes can also be defined by thepoles λp of the scattering matrix. Then, the poles are the solu-tions of the equation

det{S-1(λp)}= 0 (4)

If we extract the singular part of S(λ) corresponding to theeigenmodes of the structure, we can write S(λ) in an analyticalform as

S(λ) = ∑p

Rp—λ-λp

+Sh(λ) (5)

This is a generalized Laurent series, where Rp are the residuesassociated with each pole λp. One notes that λp = λp

R+ iλpI are

complex numbers, and that the imaginary part λpI can be linkedto lifetimes of the eigenmodes. The divergent behaviour of thefractional terms underlines the resonant character of the eigen-modes (one talks then about the resonant terms of the S ma-trix). Sh(λ) is the holomorphic part of S(λ) which correspondsto purely non-resonant processes. Ideally, in the vicinity of aspecific pole, the holomorphic part can be assumed to be al-most constant. Then, as the amplitude s(λ) of a specific dif-

fraction order is related to an element of the scattering matrix,the intensity I�Is(λ)I2 of an order can be easily derived. Onethen obtains

IF = I~0(λ-λRz—(λ-λRp)2+λ I

p2

)2+λ Iz

2

— (6)

where λz= λzR+ iλz

I is the complex zero of s(λ) in the vicin-ity of λp in the complex plan. This function leads to a typicalasymmetrical profile, a Fano profile, related to a Fano reso-nance. A typical Fano profile appears in Fig.5 (black curve).Now, if the holomorphic part is considered as negligible, onegets the following expression of the amplitude

IBR= I0—1

(λ-λRp)2+λ Ip

2— (7)

which gives a simple Lorentzian profile, related to a Breit-Wigner resonance (see for instance the red curve of Fig.5). Asshown in Fig.5, for a given eigenmode (i.e. λp imposed), thevalue of λ for which IBR is a maximum (i.e. λ = λp

R ) does notcorrespond to the location of IF maximum or minimum (i.e.(a) or (b) in Fig.5).

In order to illustrate the physical mechanisms responsiblefor the behaviour described by the scattering matrix formal-ism, we represent the corresponding processes involved.

In Fig. 6, circles A and B represent diffracting elements e.g.,holes. So an incident homogeneous wave i (in red) diffracts inA and generates a nonhomogeneous diffraction order (bluedashed line). Such an order is coupled with an eigenmode,which is characterized by a complex wavelength λp. It then be-comes possible to excite this eigenmode, which leads to a feed-back reaction on the nonhomogeneous order (one can talkthen about a resonant order). This process is related to the res-onant term of the scattering matrix. In Fig.5, the red curve isthe electromagnetic field amplitude calculated for such a non-homogeneous resonant order. The resonant diffraction order •••


features

� Fig. 4: Diagrams of slab modes. The electrostatic potential V of modesis plotted as a function of the location in the slab. Above: antisymmetric andsymmetric modes. Below: pure surface mode, with each interface decou-pled from each other.

� Fig. 3: Com-puted opticaltransmission of a thin goldfilm (200 nm)deposited onglass. 100 nm is set for theradius of eachhole and the lat-tice parameteris taken to be a = 700 nm.

••• diffracts then in B, and generates a contribution to a ho-mogenous diffraction order, in our case the specular beam.Thus one can ideally expect to observe a resonant profile, i.e.Lorentzian-like, for the amplitude sr of the homogenous spec-ular diffraction order that appears in B. Nevertheless, it is alsonecessary to account for nonresonant diffraction processes re-lated to the holomorphic term Sh(λ). So the incident wave i,generates also a homogeneous order Sh (in green), which ap-pears as another contribution to the specular beam. So, as oneobserves the specular transmitted beam, one observes the sumof two rates, i.e. sr+ sh. So, the Fano profiles of the transmission(but also of the reflection) result from a superposition of reso-nant and nonresonant contributions to the observed diffrac-tion order [10-11]. As a result, in Fig.5 one can observe thecalculated transmission (black curve) in comparison with thenon-homogeneous resonant order amplitude (red curve).

Consequences and observationsAs shown in Fig.3, the observed optical transmission throughsubwavelength hole arrays, exhibits a set of peaks and dips. Fol-lowing the previously described approach, we have shown inrecent works that the transmission spectrum is better depictedas a series of Fano profiles (see Fig.3 and Fig.5) [10,13,14].These recognizable line shapes result from the interference ofnonresonant transfers with resonant transfers, which involvethe eigenmodes film, and evanescent diffraction orders. Wecould then point out that each transmission peak-dip pair isnothing other than a Fano profile. The appearance of an assy-metric Fano line shape does not necessarily locate the peak orthe dip at the eigenmode resonance (as shown in Fig.5), con-trary to what has been suggested in some earlier works [7,8].However, we have shown that the existence of the eigenmodesis a condition for the presence of the Fano line shapes. It can benoted that recent results by Genet et al confirm this descrip-tion [11]. As a further outcome of this work it became clear thatthis kind of transmission spectrum, with its Fano line shapes,

could also be obtained in a large context and that the genericconcept of eigenmodes must substitute that of SPs only. Theseresults implied two important ideas. First, according to the con-tributions of resonant and nonresonant processes in the Fanoprofile, eigenmodes can be associated with wavelengths closerto the peaks or to the dips. Second, it is possible to obtain trans-mission curves similar to those of metal films, by substitutingSPs with guided modes or other types of polaritons. Many ex-amples can be found, including highly refractive materialsdefining guided modes or ionic crystals in the restrahlen banddefining phonon polaritons. For instance, we have reportedsimulations of a device, which consists of arrays of subwave-length cylindrical holes in a tungsten layer deposited on a glasssubstrate [13]. Tungsten becomes dielectric on a restricted do-main of wavelength in the range 240–920 nm, i.e., the real partof its permittivity becomes positive. So, whereas plasmons can-not exist, we show that the transmission pattern is similar tothat obtained in the case of a metallic film. Indeed, it was alsoshown that in this case, instead of SPs, guided modes are ex-cited [13]. Nevertheless, these modes give less intense fieldsthan those achieved with SPs. In this way the benefit for thetransmission is less important than with SPs modes. In thesame way, it was also demonstrated that the transmission pro-file of a chromium film, in the restricted wavelength domain(1112–1292 nm), where the dielectric constant is positive,should involve eigenmodes, which are not SPs or guided modes.In this case, SPs can be substituted by Brewster-Zennek (BZ)modes [14]. Lastly it should be noted that in both cases, thesetheoretical computations are supported by experimental results.

As a consequence, it appears that bounded modes can act asa mediator of the optical transmission in bidimensional sub-wavelength hole arrays. Although it appears that the mecha-nisms involved in these devices have been implicitly known fora long time, it seemed important to us to recall it. Indeed, animportant engine of the technological development is also the


features

� Fig. 6: Diagrammatic representation of the processes responsible for thebehaviour of the transmission properties.

� Fig. 5: Computed optical transmission (black curve) and amplitude of therelated resonant order (red curve) of a thin fictitious film (the permittivity isequal to i1) deposited on glass. 500 nm is set for the radius of each holeand the lattice parameter is taken to be a = 1000 nm. A discontinuityappears at the Rayleigh wavelength λRay.

improvement of our theoretical knowledge of the phenomenawith which we are confronted. �

About the authorsMichaël Sarrazin is scientific collaborator at the University ofNamur (FUNDP), Belgium. He also teaches physical sciences atthe Cours Notre-Dame des Anges in Belfort, France. He re-ceived his PhD in physics from the University of Franche-Comté, France (2002). His main research interests are intheoretical study of electrodynamics in various situations anddimensionalities. Jean-Pol Vigneron is professor of Physics at the University ofNamur (FUNDP), Belgium. His main interest lies in the field ofsurface electrodynamics and photonics, with an emphasize onthe inverse engineering of natural photonic nanostructures.

References[1] R. W. Wood, Phil. Mag. 4, 396 (1902). R. W. Wood, Phys. Rev. 48, 928

(1935)

[2] Lord Rayleigh, Proc. Roy. Soc. (London) A79, 399 (1907)

[3] U. Fano, Ann. Phys. 32, 393 (1938). U. Fano, Phys. Rev. 124, 1866(1961)

[4] A. Hessel and A.A. Oliner, Appl. Opt. 4, 1275 (1965)

[5] D. Maystre, M. Nevière, J. Opt. 8, 165 (1977)

[6] M. Nevière, D. Maystre, and P. Vincent, J. Opt. 8, 231 (1977)

[7] T.W. Ebbesen, H.J. Lezec, H.F. Ghaemi, T. Thio, and P.A. Wolff, NatureLondon 391, 667 (1998)

[8] Q. Cao and Ph. Lalanne, Phys. Rev. Lett. 88, 057403 (2002)

[9] H. J. Lezec and T. Thio, Opt. Express 12, 3629 (2004)

[10] M. Sarrazin, J-P. Vigneron, J-M. Vigoureux, Phys. Rev. B 67, 085415(2003)

[11] C. Genet, M. P. van Exter, and J. P. Woerdman, Opt. Commun. 225,331 (2003)

[12] W.L. Barnes, W.A. Murray, J. Dintinger, E. Devaux, and T.W. Ebbesen,Phys. Rev. Lett. 92, 107401 (2004)

[13] M. Sarrazin, J-P. Vigneron, Phys. Rev. E 68, 016603 (2003)

[14] M. Sarrazin, J-P. Vigneron, Phys. Rev. B 71, 075404 (2005)

europhysicsnews

features

The setting sun plays a few tricks that any physicist willappreciate. One of these is well known: the sun appears

unusually red when setting. It is the 1/λ4 dependence ofRayleigh scattering which selectively removes the blue end of thespectrum from the transmitted light.

Less well-known is the fact that the sun is not where it seemsto be, during sunset. In fact, it may be behind the horizon whilewe still see it. We are not talking here about the finite speed oflight, which makes us see the sun about 8 minutes late. We aretalking about the refraction of the sunlight due to the verticalgradient in the index of refraction, which in turn is caused bythe density gradient. If we ignore temperature gradients for asecond, the density decreases with height due to the decreasingatmospheric pressure, by a little over 1% for every 100 meter,i.e., n-1(dn/dz) ≈ 1x10-4. As a result, the light rays are bent down-ward, in the direction of the earth’s curvature. This may be seenas the inverse of the well-known ‘highway mirage’, the apparentpools lying across the pavement when the sun shines.

Granted: temperature effects can be larger than the baro-metric pressure effects, which is easily seen if we realize that, forconstant pressure, we have n-1(dn/dz) = -T-1(dT/dz). But letus look at what happens if temperature gradients are negligi-ble, or – even nicer – if temperature increases with altitude.Then the temperature effect – if any - adds to the baro-metric pressure effect. Now the light rays tend to fol-low the earth’s curvature, which makes us see the sunjust after sunset. This effect occurs both at sunrise

and at sunset, and adds an extra 5 minutes of daylight to eachday. Note that the finite-speed effect mentioned above does notdo that; it just gives an 8-minute offset throughout the day.

Since bending of light rays in the atmosphere is stronger forlower-lying rays, there is a second phenomenon: the sun ap-pears to be flattened by about 10%. The fact that we do not al-ways notice this is due to the competing effect of temperature.

Finally, there is the somewhat mysterious ‘green flash’ thatpeople sometimes observe at the moment of sunset. It lasts onlyfor a few seconds, and requires somewhat favourable atmos-pheric conditions. Why green, and why only for a few seconds?There are a few things here that we have to combine. First therefraction, which makes us see sunlight after the actual sunset.Due to dispersion this effect is strongest for the blue end of thevisible spectrum. This means that we expect blue to be visiblelongest, while the red end of the spectrum has long disap-

peared. But blue light is almost absent in the settingsun, as seen above. The result is that the last flashof sunlight is dominated by green.

The green flash: a last good-bye from thesetting sun. But at least a good-bye in style. �

Fun with the setting sun [DOI: 10.1051/EPN:2007012]

L.J.F. (Jo) Hermans, Leiden University • The Netherlands • E-mail: [email protected]


Who, in his childhood, was not astonished to see the linearmomentum of a gentle breeze or a waterfall transformed

to the angular momentum of a wheel? What wind and waterhave done for centuries, a thermal current can do today. A beamof absorbed radiation creates heat, a thermal current, which isdirectly and continuously transformed to a rotating wheel.

The development of high field magnets, such as those madeof NdFeB (Neodymium-Iron-Boron) enables to use heat to ro-tate a wheel at room temperature. Such magnets, formed assmall cylindrical discs, can be levitated in a magnetic dipolefield [1,2]. If locally heated at the upper surface the levitatedmagnet starts to oscillate with increasing amplitude until it ro-tates like a wheel. It will do so as long as the heat flow is main-tained. Radiation energy is transformed, at least partially, intorotational energy. The thermo-magnetic wheel is an exampleof self-exciting, oscillatory periodic motion that ends in a sta-ble cycle limit, where there is a balance between energy inputand dissipation [3].The amazing effect is illustrated by a video,which can be obtained from the author upon request.

ExperimentThe experimental setup is sketched in Fig. 1 and pictured inFig. 2. The lifting magnets [B] are blocks or cylinders of NdFeB,axially magnetized, with maximal field strength of B=0.5 T atthe surface and about 6 mT at the position of the floating mag-net. The bismuth discs are cylindrical, approximately 30 mmin diameter and 13 mm in thickness. The geometry of thesemagnets and discs is not crucial. The floating magnet gener-ates a maximal magnetic field of 0.5 T and is 8 mm in diame-ter, 3 mm thick, and weighs 1.2g.

Stable magnetic levitation is excluded by the well-knownEarnshaw theorem [4]. However, by placing a diamagnetic ma-terial, such as bismuth, at the right position, a three-dimen-sional potential well is created. In this well, stable levitation infree space is possible. The small magnet is supported by care-fully adjusting the lifting magnets so as to create a free-floatingposition between the bismuth plates.

What is seen?The motion of the magnet is recorded with both a video cam-era and a LASER beam. When the beam is reflected from amarked surface of the magnet, it hits a photocell. When thelight is on, i.e. when a continuous heat input is present, motionstarts from the equilibrium position, and the magnet behaveslike a self-excited oscillator. The oscillation begins at an angu-lar frequency ω(t) of about 2 rad/s. After several oscillations,ω(t) decreases to a minimum, due to the anharmonic motion atlarge angles. The motion then rapidly changes into a perma-nent steady-state rotation. Fig. 4 illustrates two typical exam-ples of this motion.

A physical model The thermo-magnetic wheel is modelled as a circular disc inharmonic rotational motion. The disc is suspended at a point,which we denote as magnetic suspension point [msp]. Thispoint does not generally coincide with the centre of mass [cm].Fig. 5 shows the system in the x-y plane normal to the z-axis.

The magnetic levitation force, which we denote by Fm,matches the weight Fw and has an action point, which lieshigher than the gravitational centre of mass [cm]. The actionpoint is not at [cm] because there are always slight inhomo-geneities of the total magnetisation M with respect to the cylin-drical geometry of the magnet. The magnet would be unstablein the z-axis direction without the Bi plates, and would quicklybe pulled to one of the lifting magnets. With the diamagneticbismuth plates, however, there are repulsive forces acting on thelevitated magnet that obey an inverse 4th power law with dis-tance, leading to a stable equilibrium position.

If gently displaced from its equilibrium position, the magnetwill oscillate about the magnetic suspension point [msp] inboth the x and y directions, horizontally and vertically. Thismeans that the magnet is suspended in the magnetic field in


features

A thermo-magnetic wheel [DOI: 10.1051/EPN:2007013]

Claudio Palmy,Alpine Institute of Physics Stuls (APHIS) • CH-7482 Stugl/Stuls • Switzerland

� Fig. 2: The lifting magnets are quadratic blocks, 25x25 mm and 12 mmthick. They can be moved on a track. The floating magnet lies between thebismuth plates and floats without touching them. A 12V spotlight is theradiation source.

� Fig. 1: The parts of the system are: The lifting magnets [B], the bismuthdiscs [Bi], the floating magnet [fm] and a radiation beam [rb].

every direction. Without radiation the magnetic suspensionpoint is fixed. The motion is given by the differential equationfor harmonic motion:

Iφ••

+Rφ•

+D*φ = 0 (1)

where the •

stands for the time derivative, I is the moment of in-ertia, R the coefficient of the friction torque Rφ

•(t), and D* = m

g D, the coefficient of the restoring torque (D*φ), and D is thedistance between the action points of the magnetic and gravi-tational forces. φ is the angle of rotation shown in Fig. 5.

We measured the natural angular frequency of the systemto be ωo = 2.1 rad.s-1, and we estimated a mean oscillation decaytime to be approximately τosc = 50s, so we can calculate all rel-evant oscillation parameters, R, D* and D. With I = 9.6 10-9

kgm2 we get: D=ωo2 I/mg = ωo

2 r 2/2g = 3.6 10-6m, D* = ωo2 I = 4.2

10-8 Nm/rad, and R = I/τosc. = 1.9 10-10 Nms/rad, where m = 1.2gis the magnet’s mass and g is the gravitational acceleration.

What is the reason for the self-excitation?This thermo-magnetic wheel shows a startling behaviour whena heat flow is applied from above. The magnet starts to oscillatefrom rest with increasing amplitude. Some immediate ques-tions are: Why does the amplitude of this oscillation increase?What is the source of the oscillation energy? What is the en-ergy-transfer mechanism, and what is the equation of motionfor the angular displacement φ between 0 and 2π?

Let us consider the equilibrium position for the disc magnetsketched in Fig. 6. The potential energy U of the floating mag-net with a magnetic dipole moment M in the magnetic field Bof the lifter magnet is [2]:

U = M.B+mgy = -MB+mgy (2)

where mgy is the gravitational energy. In the absence of heat the floating magnet will align with the field B. There is no magnetictorque. The resulting net force Fnet = -grad U on the magnet is zero.

When the light is on, the upper magnet’s surface is heated,and the local magnetisation decreases. Consequently the mag-netic suspension point [msp] sinks gradually for about 1mmto a new equilibrium position inside the inhomogeneous liftingB-field. This shift of the [msp] generally leads to a misalign-ment of M and B, and a torque τ appears.

The heated region, which got a “thermal kick” at t = 0, movesaway from the direct heat influx area and cools exponentially.Denoting the local magnetisation, i.e. the part of M which isaffected by the thermal flow, as MT, we have: MT = Moexp(-t/t*)[5]. The phenomenological constant t* describes the tempera-ture decline of the magnetic dipole MT, as sketched in Fig. 6.Mo is the difference of the magnetic dipole between the initialand final states when the magnet rotates permanently, Mo =Minitial – Mfinal.

Since M and B are no longer aligned, a torque τ = MT x Bacts on the magnet in addition to the gravitational one. Thistorque resembles the one observed in the gyro-magnetic ex-periment of Einstein and de Haas in 1915 [6]. There, the changein direction of the magnetic dipole M also causes a correspon-ding change in the angular momentum of a ferromagnet.Therefore there is a torque τ = MTxy Bxy sin(φ) about the z-axis,where MTxy and Bxy are the respective magnetic vectors in thex-y plane, and φ = t*φ

•is the angle between them. We have

τ = A sin(t*φ•

) (3)

where A and t* are constants, chosen by best fit to the numer-ical solutions for the magnet’s motion. The additional torque τdoes not depend on time explicitly, but it acts on the pendulum •••


features

� Fig. 3: Magnified section of the floating NdFeB disc magnet betweenthe diamagnetic bismuth plates. The levitated magnet is 8mm in diameterand 3 mm in thickness.

� Fig. 4: The angular frequency ω(t) is around 2 rad/s for the oscillationsequence. The angular speed ω(t) for the rotation is drawn as function of timefor two heat intensities.

••• just as the gravitational restoring torque does. If we add thistorque to Equation (1) leaving, however, the small angle ap-proximation, we get:

Iφ••

+Rφ•

+D*sin(φ) = A sin(t*φ•

) (4)

Equation (4) is a homogeneous, non-linear differential equa-tion with no external independent function of time. We solveEquation (4) by numerical approximation using MATHE-MATICA with the numerical values ωo

2 = D*/I = 4.4 [s-2], φ[0]=0, φ

•[0] = 0.2 [s-1], R/I= 0.02 [s-1], A/I = 0.9 [s-2] and t*=

0.25 [s]:

φ••

+0.02φ•

+4.4 sin(φ) = 0.9 sin(0.25 φ•

) (5)

In Fig. 7 we plot the angular velocity ω (t) = φ•

(t) obtainedafter numerical differentiation of φ(t). The angular displace-ment φ(t) is the numerical solution of equation (5). Fig. 7 is aquantitative correct picture of the observed self-excited oscil-lation and subsequent rotation. An oscillation is superposed onthe rotation; it is due to the slight eccentricity of the thermo-magnetic wheel. Also, equation (5) gives solutions for the caseswhen light does not hit the magnet vertically, but is displacedby π/2 or π, i.e. when φtorque = φ +/-π/2 or φtorque = φ + π; thesecases are experimentally confirmed. This mechanism of self-excitation by a continuous thermal flow resembles very muchthe wind-driven oscillations observed in 1940 at the TacomaNarrows bridge in the USA [7].

After the transient self-excited oscillations end, the magnetcontinues in rotation with a constant mean angular speed ω(t).The temperature of the magnet reaches a constant value, T∞,and the thermal gradient goes away. A new, dynamically stablestate is reached. The active magneto caloric forces, which initi-ated the oscillation, become a small torque. This torque matchesthe minute external torque due to air drag and, perhaps, eddycurrent dissipation.

Is the thermo-magnetic wheel a heat engine?Yes, because this simple room temperature experiment trans-forms radiation heat directly to rotational motion of a wheel.This amazing effect can be described completely within theframework of a thermo-magnetic self-excitation mechanism.The thermal flow through the levitated magnet provides energyto sustain the oscillation and subsequent rotation. However, itspotential as a source of mechanical power is presently not veryattractive. A rotating magnetic field is, in principle, a Faraday discgenerator. We estimate the heat energy flowing per secondthrough the magnet to be around 10 mJ, while the rotating en-ergy of the magnet is less than 1µJ. This minute rotational en-ergy cannot compete with solar-based energy converters.

This analysis describes an angular pendulum that transfersenergy between radiation-, magnetic- and gravitational fields.The model adequately explains the observations. This ther-mally driven "heat wheel" combines gravity, magnetism andelectromagnetism, as well as thermodynamics and mechanics,in a simple demonstration that is unique and unusual. �

About the author:Claudio Palmy received his Ph.D. in superconductivity fromthe Swiss Federal Institute of Technology Zürich, ETHZ, in1970. After a short time as a visiting professor in Sao Paulo,Brasil, he became a physics professor at NTB, Interstate Uni-versity of Applied Sciences of Technology Buchs, Switzerland.Since 2004 he has worked on magnetism at the Alpine Instituteof Physics Stuls, Switzerland. E-mail: [email protected]

AcknowledgementsThe author thanks Jean-Pierre Blaser, Peter W. Egolf and RenéMonnier for many helpful discussions. He also thanks RogerAlig for careful and thorough reviewing of the paper.

References

[1] M. V. Berry and A. K. Geim, Eur. J. Phys. 18, 307 (1997) andwww.hfml.sci.kun.nl/hfml/levitate.html

[2] M. D. Simon, L.O.Heflinger and A.K.Geim, Am. J. Phys. 69, 702 (2001)

[3] C. Palmy, Eur. J. Phys. 27, 1289-1297 (2006)

[4] S. Earnshaw, Trans. Cambridge Philos. Soc. 7, 97 (1842)

[5] J. P. Blaser, personal communication (2006)

[6] A. Einstein, W. J. de Haas, Verhandl deutschen physikalischen Ges 17,152 (1915)

[7] K. Y. Billah and R. H. Scalan, Am. J. Phys. 59, 118 (1991)


� Fig. 7: Calculatedangular velocity ϕ• (t)= ω(t) for a levitatedmagnet in a vertical(y-axis) temperaturegradient at roomtemperature.

� Fig. 6: The magnet’s position afterreceiving a “thermal kick” at t = 0 atthe position ϕ = 0. The positionangle ϕ is t*ϕ• , where t* is the timeconstant of the local exponentialtemperature decay. The magneticdipole MT and the levitating B-fieldare shown. The static forces Fm and Fw

are omitted.

� Fig. 5: Levitated magnet as an oscillat-ing disc. [D] is the distance between

the centre of mass [cm] and the mag-netic suspension point [msp] whileϕ denotes the rotational angle meas-ured relative to the equilibrium axis

defined by msp and cm. The z-axis isdirected out of the XY plane. The static

forces Fm and Fw are shown.

features

europh ysicsnews · 3page31 page23 page 18 number3• volume38 • 3 contents europh ysicsnews...

Documents