THE NEWS MAGAZINE OF THE ESRF - ALSO AT http://www.esrf.fr

E U R O P E A N S Y N C H R O T R O N R A D I A T I O N F A C I L I T YAPRIL 1998 N° 30

ISS

N 1

011-

9310

INTERNATIONAL CONFERENCE

Highlights in X-Ray Synchrotron Radiation Research

17-21 November 1997Grenoble, France (see page 3)

IN THIS NEWSLETTERReports

from workshops on Magnetism (page 13), X-ray Imaging (page 17) and High Pressure (page 25).

A reviewof the Collaborating Research Groups at the ESRF (page 32)

ESRF - APRIL 1998

2

NEWSLETTER

IN BRIEF

MAGNETISM

WORKSHOP

X-RAY IMAGING

WORKSHOP

HIGH PRESSURE

WORKSHOP

CRGREPORTS

RESEARCH &DEVELOPMENT

EVENTS

The Conference on «Highlights in X-Ray Synchrotron Radiation Research», PAGE 3, M. Altarelli.Biology Symposium, PAGE 6, P. Lindley.Applications for Beam Time Arrive from Labs World-Wide, PAGE 7, R. Mason.«Powder Diffraction at Synchrotrons», PAGE 8, A. Fitch/Å. Kvick.«Ultra-Fast Kinetics of Molecular Assemblies», PAGE 8, M. Wulff.ESRF at «Physique 97» in Paris, PAGE 9, J. Doucet.27th and 28th Council Meetings, PAGE 10, K. Witte.Georg von Klitzing New Chairman of the ESRF Council, PAGE 10.Open Days, PAGE 11.Theatrical Itinerary «Lumières», PAGE 11.

X-Ray Scattering and Magnetism, PAGE 13, C. Vettier and G. H. Lander.Detection of Orbital Order in Transition Metal Oxides by Resonant X-Ray Scattering, PAGE 14, M. Altarelli.Probe Coherence Volumes and the Interpretation of Scattering Experiments, PAGE 14, N. Bernhoeft, A. Hiess,A. Stunault, S. Langridge, D. Wermeille, C. Vettier, G. H. Lander, M. Huth, M. Jordan and H. Adrian.

Meetings on X-Ray Imaging, PAGE 17, J. Baruchel.Combining Diffraction Topography and Phase Imaging, PAGE 18, J. Baruchel, P. Cloetens, M. Drakopoulos,L. Mancini , P. Rejmánková-Pernot, A. Snigirev, I. Snigireva, A. Souvorov.In Situ Damage Assessment in Micro-Heterogeneous Materials using High Resolution X-Ray Tomography,PAGE 20, J-Y. Buffiere, E. Maire, P. Cloetens, G. Peix, M. Salomé, J. Baruchel.Bragg Diffraction Imaging of Magnetic Crystals: New Results from Novel Beams, PAGE 22, C. Medrano,I. Matsouli, J. I. Espeso, V. V. Kvardakov, M. Schlenker, J. Baruchel.

High Pressure Workshop, PAGE 25, D. Häusermann.Laser Heating under Pressure: A Brilliant Journey Towards the Centre of the Earth, PAGE 26, G. Fiquet,D. Andrault, A. Dewaele, T. Charpin, F. Visocekas, D. Häusermann, M. Kunz and T. LeBihan.Density Measurements of Liquid Iron Alloys at High Pressures: Towards a Better Understanding of the Planets,PAGE 28, C. Sanloup, I. Martinez, F. Guyot, P. Gillet, G. Fiquet, M. Mezouar, D. Häusermann and T. LeBihan.

Collaborating Research Groups: a Review, PAGE 32, I. Kilvington.SNBL (Swiss-Norwegian Beamline) on BM1, PAGE 34.

- A New Technique: Physically Estimated Triplet Phases, PAGE 34, R.H. Mathiesen, F. Mo, A. Eikenes,T. Nyborg and H.B. Larsen.- EXAFS: Hydride Treatment Catalysts, PAGE 36, T. Shido, R. Prins and H.P. Weber.

D2AM (French Beamline) on BM2: Protein Crystallography and Materials Science, PAGE 38.- The High Temperature Project, PAGE 39, C. Landron, L. Hennet, J.P. Coutures, M Gailhanou, M. Gramondand J.F. Berar.

GRAAL on BM7: Gamma-Ray Beam, PAGE 41, C. Schaerf and D. Rebreyend.GILDA (Italian Beamline) on BM8, PAGE 42, F. d'Acapito, S. Colonna, S. Pascarelli, G. Antonioli, A. Balerna,A. Bazzini, F. Boscherini, F. Campolungo, G. Chini, G. Dalba, G. I. Davoli, P. Fornasini, R. Graziola, G. Licheri,C. Meneghini, F. Rocca, L. Sangiorgio, V. Sciarra, V. Tullio and S. Mobilio.ROBL (German Beamline) on BM20: Structural and Radiochemical Investigations, PAGE 45, W. Matz, N. Schell,H. Funke and G. Bernhard.SpLine (Spanish Beamline) on BM25, PAGE 46.DUBBLE (Dutch-Belgian Beamline) on BM26, PAGE 47, M. Oversluizen, P. Goedtkindt, W. Bras,D. Detollenaere, R. van Tol and W. de Jeu, with support from I. Cerjak, D. Glastra van Loon, P. Lassing,B. Munneke, J. Derks, M. Konijnenburg, M. Borsboom, H. Alberda, R. Bakker and F. Udo.XMAS (UK Beamline) on BM28: X-ray Magnetic and High-Resolution Scattering, PAGE 47.FIP (French Beamline)on BM30: Investigation of Protein Structures, PAGE 49, M. Roth.IF (French Beamline) on BM32: Interfaces, PAGE 49, G. Renaud.

- Environmental Science, PAGE 51, G. Sarret, A. Manceau and J.L. Hazemann.

Real-time Phase-Contrast Imaging, PAGE 53, P. Cloetens, J.I. Espeso and F. Sever.Mechanically Induced Influences on the Storage Ring Beam, PAGE 54, D. Martin and L. Farvacque.

Open Days, 14-15 March 1998.Theatrical Itinerary «Lumières», from 20 to 31 March 1998.

Photography by:

C. Argoud, S. Bertini,S. Berard and T. Le Bihan.

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

3

In 1947, the first report of directvisual observation of synchrotronradiation from a 70 MeV machine inthe General Electric ResearchLaboratories appeared in the PhysicalReview. To celebrate the fiftiethanniversary of this momentuousevent, and to provide a timely reviewof the status of the field after the startof operation of third generationsources around the world, ESRForganized an international conferenceon «Highlights in X-Ray SynchrotronRadiation Research». The conferencewas held in Grenoble from 17 to20 November 1997, and wasaccompanied by a symposium onStructural Biology and followed bythe ESRF Users’ Meeting and by thegeneral meeting of the EuropeanSynchrotron Radiation Society.Specialized workshops were alsoorganized immediately before andafter the conference. More than 500attendees convened and provided aresponsive audience who animateddiscussion sessions throughout themeeting.

The format chosen for theconference was to have a series ofplenary invited talks, at least 35minutes long, most of them groupedso as to provide a broad overview of ahighlighted field of science. Livelyposter sessions, complementing theplenary talks, concluded three of thefour conference days.

On the first day, after a shortwelcome from Y. Petroff, Chairman ofthe Organizing Committee, B.Batterman of Cornell University took upthe challenge of taking the audience ona trip through fifty years of synchrotronradiation in just about 40 minutes. Thiswas no easy task, and the livelydiscussion at the end of his presentationincluded questions challenging hisomission of one or another contributionto the history of the field. In thefollowing presentation J. M. Filhol,ESRF Machine Director, reviewedthe performance of 3rd generationsources and gave an outlook on whatcould be expected from the4th generation sources, which are at

present on the drawing board.The x-ray investigation of magnetic

systems has been undergoing aqualitative and quantitative explosion inrecent years, thanks to the wavelengthselection and polarization propertiesallowed by synchrotron radiation. Theearly part of the conference featuredplenary talks devoted to an in-depthdiscussion of the status and theperspective of magnetic studies withsynchrotron radiation.

After a review of the theoreticalfoundations of the different techniquesby M. Blume of Brookhaven, talks byC.T. Chen (SRRC, Taiwan), P. Carra andJ. Goedkoop (ESRF) discussed magneticcircular dichroism, and the experimental

and theoretical breakthroughs whichmade this technique so popular anduseful for the study of ferromagneticsystems. E. Kisker (DüsseldorfUniversity) discussed methods based onphotoelectron detection, in particularthe imaging and element-sensitivecapabilities. G. Sawatzky (GroningenUniversity) discussed an area in whichsynchrotron radiation shows greatpromise of becoming a very importanttechnique, but has not yet experimentallyproven it: orbital order in transition metaloxides, where the orbital degeneracy ofd-electrons can be removed by formationof a superlattice of orbital orientations,with dramatic effects on magneticstructure and excitations.

THE CONFERENCE ON

«HIGHLIGHTS IN X-RAY SYNCHROTRON RADIATION RESEARCH»

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

4

On Tuesday morning, three more talksdevoted to magnetic studies, specificallyby magnetic scattering, were presented.D. Gibbs (Brookhaven) reviewed workon many different systems, and alsoemphasized the emerging field of surfacemagnetic scattering. This technique wasdiscussed in detail by S. Ferrer of ESRFwho showed the recent results onthe Co/Pt system obtained at ID3.M.J. Cooper (Warwick University)described recent Compton scatteringstudies of the spin density.

In the next group of talks, S.K. Sinha(APS, Argonne) and G. Gruebel (ESRF)discussed and exemplified the use of thepartially coherent character of undulatorx-ray radiation in the scattering studies ofdynamics and fluctuations in condensedmatter. This is certainly one of the mostnovel aspects of third generation sources,and, in spite of the many beautiful resultspresented, one which has yet to deliver itsfull promise of exciting new knowledge.In another field which has known a fastdevelopment of synchrotron radiationstudies, soft-condensed matter, J. Daillant(Saclay) discussed grazing incidence

studies of liquid surfaces and interfaces,and A.J. Ryan (University of Sheffield)and M. Stamm (MPI Mainz) reviewedrecent applications to polymer science.The last two plenary talks of the day weredevoted to inelastic scattering with ultra-high energy resolution. F. Sette (ESRF)described the spectacular results of ID16on the dispersion of collective excitationsof liquids and amorphous systems,measured with meV resolution inpreviously inaccessible regions of theenergy-momentum plane. E. Burkel(University of Rostock) presented resultson crystalline solids, including high-pressure investigations of phonons in He.

On Wednesday, W. Hendrickson ofColumbia University described progressin the investigation of biomolecularstructures using the MAD method, whichis taking place thanks to dedicatedbeamlines and to the quality of undulatorradiation. J. Trümper (MPI, Garching)delivered a fascinating presentation of x-ray astrophysics, with splendid pictures ofcelestial bodies from our small moon toimmense galactic clusters, and showedhow progress in x-ray instrumentation is

widening and deepening our view of theUniverse.

Coming down to earthly applicationsof synchrotron radiation, J. Schneider(Hasylab) described the use of very high-energy x-rays from Petra undulators toperform beautiful experiments on thecritical region surrounding the structuraltransition of SrTiO3, and to elucidate the«two length-scales» behavior induced bystructural defects in the near surfaceregions of the sample. E. Gerdau(University of Hamburg) reviewednuclear resonant scattering of synchrotronradiation, which is still delivering newand sometimes surprising applications ofthe extremely monochromatic beams itproduces, which have thereforelongitudinal coherence lengths of manymeters.

The remaining talks of the dayexplored theoretical and experimentaladvances in high-pressure physics. On thetheoretical front R. Martin (University ofIllinois) described the present knowledgeof the phase diagram of hydrogen, and thelong-standing effort to pinpoint and,hopefully, observe its insulator-to-metal

USERS’MEETING

The annual Users' Meeting tookplace on 21 November 1997, followingthe four-day SR50 Conference. It wasorganized around one plenary sessionand three mini-workshops on X-RayImaging, Time-Resolved Scatteringand Absorption, and also IndustrialApplications of Synchrotron Radiation.

The «Young Scientist of the Year»award, which is given every year bythe Users' Organization for outstandingwork done at the ESRF by a scientist35 years of age or younger, was offeredthis time to M. Thoms (from ErlangenUniversity) for the development of anew image plate detector, an order of

magnitude faster than other detectorsof this kind.

The new Council of the Users'Organization has been elected forthe next two years: P. Armand,F. Boscherini, J. Cockroft, M. Frey,A. Kaprolat, D. Nicholson, withM. Cooper as chairman.

M. Thoms receiveshis award fromthe handsof S. Pascarelli,Chairwomanof the Users’Organization.

The new Councilof the Users’

Organization.

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

5

transition, with its many implications forfundamental solid state physics as wellas planet physics; and M. Parrinello(MPI Stuttgart) reported ab initiomolecular dynamics simulations ofwater and ice, where the hydrogen bondsplay a fundamental role. In particular, hediscussed the phase transformations ofice at high pressure. On the experimentalside, P. Loubeyre (University of ParisVI) described the large impact of 3rdgeneration synchrotron radiation sourceson single crystal structural studies oflight elements like hydrogen and heliumat Mbar pressures; H.K. Mao (CarnegieInstitute, Washington) reported howthe simultaneous development ofsynchrotron sources and of high pressureinstrumentation such as diamondanvil cells equipped with Bewindows and improved control ofpressure and temperature are allowingthe investigation of materials underconditions approaching those in the deepcore of the earth. The same messagecame through the talk of D. Andrault(Institut de Physique du Globe, Paris),describing experiments on Fe at 2500 Ktemperatures and Mbar pressures andtheir implications for geophysics.

The final day of the conferencecomprised a presentation by S. Doniach(Stanford) on time-resolved SAXSstudies of protein folding and then aseries of talks devoted to x-ray imaging.Doniach reported how proteins, after aninitial rapid collapse to a metastablestate, approach the final configurationwith a single rate constant of order 10-100 ms, a time scale accessible to time-resolved SAXS investigations. Thetechnique allows for example to followthe evolution of the radius of gyrationalong the folding pathway.

Imaging methods are very importantbecause, among the many applications ofsynchrotron radiation, they can beparticularly important for industrialproblems and for problems of greatpublic interest such as medical imaging.The progress in imaging based onphase contrast was discussed in thepresentations by U. Bonse (Universityof Dortmund), S. Wilkins (CSIRO,Australia), A. Momose (Hitachi, Japan)and A. Snigirev (ESRF). Bonse reportedon three-dimensional tomographicimages reconstructed from the phaseshifts suffered by a beam going throughthe sample in different directions, as

measured with an interferometer, andshowed the advantages of phase contrastover absorption for biological specimenscomposed of low-Z elements. Wilkinsdiscussed the different experimentalconfigurations and the theoreticalfoundations of phase contrast imagereconstruction, with the goal offormulating the optimum conditionsunder which a particular feature of thesample can be imaged. Momosedescribed in particular some stepstowards the application of phase contrastmethods in vivo, for which minimizationof the dose is of course a crucialcondition, to be obtained by optimizingthe equipment and the selection of the x-ray wavelength. Snigirev presented abroad overview of different imaging andfocusing techniques exploiting thepartial coherence and the highcollimation of undulator beams, and theirapplication to the characterization ofsamples, of optical elements and of thecoherence properties of the beam itself.

P. Spanne (ESRF) discussed medicalimaging with hard-x-ray synchrotronradiation and the efforts underway to

exploit coherence and diffractionproperties for radiography. Finally,J. Gronkowsky (University of Warsaw)and J. Baruchel (ESRF) discussed thestatus and the challenges of topography.Gronkowsky reported on the simulationsof topographic images, where newalgorithms have led to substantialprogress in recent years. Baruchelreported on the new scope oftopographic investigation with thirdgeneration sources. In particular hereported results clarifying the growthmechanisms and the role of defects inquasi-crystals.

In summary, the Conferenceprovided the participants with anoverview not only of the recent resultsobtained by synchrotron radiationsources but also of the scientific fieldsin which context they are taking place.The talks by theoreticians in particularallowed an appreciation of themeaning and relevance of synchrotronradiation experiments in relation to theopen questions at the frontiers ofresearch.

M. Altarelli

ESRS PRIZE GIVING

On the occasion of fifty yearsof synchrotron radiation, theEuropean Synchrotron RadiationSociety (ESRS) awarded a prizeof 1500 ecus for an outstanding

contribution to synchrotron radia-tion science. The winner wasM.I. McMahon from Daresburyfor his work in the high pressurefield.

M.I. Mc Mahon, winner of the ESRS prize.

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

6

The Biology Symposium washeld on 19 and 20 November and ranin parallel to the main SR50Conference. However two plenarylectures were held in common.Firstly, W. Hendrickson (NewYork) outlined how the MultipleWavelength Anomalous Dispersiontechnique had provided a newapproach to the «phase problem» inmacromolecular crystallography,leading to the elucidation of structureswhich had not been possiblewith conventional methods. In thistechnique the ready tunability ofsynchrotron radiation was a keyfeature. Secondly, S. Doniach(Stanford) explained how small-angle x-ray scattering could be usedto study the biologically importantproblem of how polypeptide chainsfold to yield functional, three-dimensional macromolecules. Thesymposium, itself, concentratedmainly on biological macromoleculeswhose structures had been determinedby protein crystallographic methodsand comprised four themes; virusstructures, RNA/DNA complexes,membrane proteins and time-resolved

studies of reaction mechanisms.A highlight in the virus sectioninvolved a description of the bluetongue virus by D. Stuart (Oxford).This double stranded RNA virus,which is one of the most virulentamong ruminants, has a core particlewith a diameter of about 800 Å and anoverall molecular weight of 100 MDa,thus providing a severe technicalchallenge with regard to diffractiondata collection using synchrotronradiation and structure determination.Structural interme-diates in the lifecycle of icosahedral viruses weredescribed by M. Rossmann (WestLafayette) and L. Liljas (Uppsala)explained structural studies on themechanism of assembly of small RNAphages using coat protein mutants andRNA complexes. In the secondsection an undoubted highlight wasthe structure of the nucleosome coreparticle of chromatin undertaken byT. Richmond (Zürich) and colleagues.The structure shows how four pairs ofhistone fold domains are responsiblefor organising 121 base pairs ofDNA whilst extensions of the histonefold bind the termini of the DNA

superhelix and help to stabilise theoverall structure. Other speakersdescribed principles of RNA-proteininteraction (T. Steitz, Yale), how classII aminoacyl-tRNA synthetases work(S. Cusack, EMBL, Grenoble),recognition and repair of mismatchedDNA base pairs (L. Pearl, London)and crystallographic studies ofeukaryotic gene expression (S. Burley,New York). The elucidation of thestructures of membrane proteinsprovides further stern challenges forsynchrotron x-ray crystallographictechniques. J. Rosenbusch (Basel)explained how the key step ofobtaining suitable crystals, albeit ofsmall dimensions, can be achievedwith reference to the growth ofbacteriorhodopsin in a cubic lipid gelphase. N. Isaacs (Glasgow) describedthe elegantly symmetric ring structureof the light harvesting complexfrom photosynthetic bacteria whichhas enabled an understanding ofthe process of photosynthesis andT. Schirmer (Basel) described thestructures of transmembrane proteins,general porins to maltoporin, whichallow solutes and sugars (specificallymalto-oligosaccharides) respectively,to pass through outer cell membranes.The final session was devoted tostudies of reaction mechanisms usingvery fast time-resolved measurements.

I. Schlichting (Heidelberg)described the characterisation ofintermediates in the reaction cycle ofcytochrome P450(cam), M. Wulff(Grenoble) explained the developmentof the technology required to undertakesuch studies and K. Moffat (Chicago)highlighted studies on the millisecondto nanosecond time scale on hemeproteins and the photoreactive yellowprotein.

The symposium provided a firstclass example of how x-ray structuralanalysis of macromolecules usingsynchrotron radiation is making a vitalcontribution to our understandingof life processes; all the speakers areto be congratulated on the excellenceof their presentations.

P. Lindley

BIOLOGY SYMPOSIUM

K. Moffat(Chicago) inthe finalsession.

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

7

APPLICATIONS FOR BEAM TIME ARRIVE

FROM LABS WORLD-WIDE

Scientists from laboratories inEurope, the US and Japan requested atotal of 8469 shifts on beamlines atESRF for the first seven months of1998. In total, 581 applications arrivedfor the September 1997 deadline forbeam time on the 28 ESRF and 4 CRGbeamlines which will be schedulinguser experiments during this period.

The Review Committees, whosemembers are external scientific expertsin a range of fields, met at the ESRF on23 and 24 October 1997 to assess theapplications. Following discussionslasting one and a half days, 387proposals were finally selected andallocated a total of 4914 shifts or 1638days of beam time. Of these, four newlong-term projects were accepted andallocated beam time for a two-yearperiod. Figure 1 shows the resultsof these allocations per ReviewCommittee. Slightly more than 66% ofproposals were successful this reviewround, compared with an average of45% over previous rounds. The highersuccess rate this time reflects theincreasing number of beamlinesavailable, and a slightly longerscheduling period of 7 months.

The Committees commended userson the higher standard of proposals.This largely reflects the experiencewhich experimental teams areacquiring on the beamlines, and anincreasing interest in the kinds ofchallenging projects which can beundertaken given the instrumentationavailable and the specific qualities ofthe beam at the ESRF.

The beamlines most frequentlyrequested this round were ID16,for inelastic scattering; ID12B, theDragon spectrometer with polarizationselectivity; ID2, the high-brilliancebeamline, which combines applicationsin small-angle x-ray scattering andmacromolecular crystallography; ID32,for SEXAFS and standing waves, andID18, the nuclear resonance beamline.

Figure 2 shows the number ofshifts allocated compared with shiftsrequested, per scheduling period, sincethe beginning of user operation inSeptember 1994. It should be noted that

the second scheduling period each yearto date has been slightly shorter than thefirst, so that fewer shifts have beenallocated and scheduled during thesecond half of each year.

Potential users are reminded thatthe next deadline in 1998 forapplications for beam time is 1September. During the correspondingscheduling period the availablebeam time for macromolecularcrystallography will increasesignificantly, since two additionalexperimental stations on ID14 will beoperational. Further, a second beamline

dedicated to elastic scattering forelectronic excitations, ID28, will alsobegin commissioning, so that by theend of the period the full complementof 30 publicly financed beamlines willbe operating with users. The BritishCRG beamline XMaS, for magneticscattering, will accept first users fromApril 1998. Details of the beamlinesand instrumentation available, proposalforms and experimental report formsare available at the ESRF Web site:http://www.esrf.fr, or from the UserOffice.

Fig. 1: Proposals submitted and allocated beam time,per Review Committee.

Fig. 2: Total number of shifts of beam time requested and allocated,per scheduling period: September 1994 to July 1998.

1994: 1995: 1995: 1996: 1996: 1997: 1997: 1998:Sept to Dec Jan to July Aug to Dec Jan to June July to Dec Jan to June July to Dec Jan to July

Scheduling periods

Num

ber

of s

hift

s

10 000

9000

8000

7000

6000

5000

4000

3000

2000

1000

0

Shifts requested: 55 503

Shifts allocated: 23 113

2252

761

4236

1965

6634

2034

7965

3065

7660

3046

8384

3569

9903

3759

8469

4914

Chemistry Hard Hard condensed Life Methods and Soft Surfaces andcondensed matter: sciences Instrumentation condensed Interfaces

matter: Elect. Structures matter& Magn. properties

Review committees

Num

ber

of p

ropo

sals

160

140

120

100

80

60

40

20

0

Proposals received 1998-I: 581

Proposals accepted 1998-I: 387

73

46

114

73

143

83

128

112

19 16

62

3142

26

R. Mason

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

8

This workshop was held at theESRF on 11 and 12 December 1997. It was organized by A. Fitch andÅ. Kvick and featured 12 invitedspeakers from Europe and the US(H. Fuess, TU Darmstadt - A. Fitch,ESRF - D.E. Cox, BNL - D. Louër,Univ. Rennes - R.J. Cernik, Daresbury- G. Artioli, Univ. Milano - P. Norby,Stony Brook - T. Wroblewski,Hasylab - P.J. Webster, Salford -L.B. McCusker, ETH Zürich - W.I.F.David, Rutherford - A.K. Cheetham,Univ of California). The latestdevelopments in structure determinations

from high-resolution powder data,materials characterization of industrialimportance and in situ studies ofreactions and processes usingpowder techniques were discussed.A concluding discussion session

outlined the hardware and analysisdevelopment needs for the future. Theworkshop was attended by about70 people of which more than 50scientists came from outside the ESRF.

A. Fitch/Å. Kvick

«ULTRA-FAST KINETICS OF MOLECULAR ASSEMBLIES»On 15-16 January 1998 the ESRF

hosted a workshop titled «Ultra-fastkinetics of molecular assemblies» withthe aim of reviewing progress on time-resolved diffraction experiments onphotosensitive proteins and to examinewhether this technique is applicable tosmaller scale chemical systems. 15invited speakers covered fields rangingfrom protein crystallography, opticalspectroscopy to future ligth and electronsources.

K. Moffat, who pioneered ultra-fasttime-resolved diffraction on proteinswith the scientists on ID9, presentednew data of the first intermediate in thephotocycle of the yellow protein PYP.Electron density maps of the first

nanosecond intermediate show that theshaft of the light sensing chromophore iscompressed and that the stored energy isreleased in a large rotation of thechromophore. These real-time data werecontrasted by new static measurementsby L. Getzoff, who used the «cool thenactivate» approach. In this way one canstudy intermediates preceding anactivation barrier which becomesinsurmountable at low temperatures.She showed beautiful monochromaticelectron density maps at true atomicresolution. R. Hochstrasser gave aoverview of ultra fast laser spectroscopyshowing the strength and weakness ofoptical techniques. He showed thatoptical spectroscopy probes extended

electronic levels which are often poorindicators of atomic structure. Bycontrast x-rays interact with core levelsand are therefore effective structuralprobes. C. Shank from LawrenceBerkeley Laboratory showed how theygenerate 300 fs x-ray pulses at 0.4 Å by90 degree Thompson scattering betweeninfrared terawatt pulses and electronsfrom a linac. With a brilliance of2x105 photons/mm2/mrad2/s in a 15%bandwidth, it has been possible to studythe onset of disorder in a ISn crystalfrom a 100 fs temperature jump. In thefinal session about future x-ray ligthsources, B. Wiik from Desy describedthe TESLA project, a proposal toconstruct two 16 km linacs running at250 GeV, one for electrons and one forpositrons. In the 1 x 1 nm2 collisionpoint, the high energy physics hope toproduce the energy density a picosecondafter the big bang! The instrument willallow high-energy physicists to studythe top quark and to search for the stilltheoretical Higgs Boson, the carrier ofmass. The interest to the x-raycommunity is the free electron sidestation. B. Wiik showed that a freeelectron laser will produce 100 fscoherent x-rays with 1012 photons perbunch, 104-105 times the present level atthe ESRF.

M. Wulff

«POWDER DIFFRACTION AT

SYNCHROTRONS»

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

9

The ESRF is well-known as apublic research institute but probablyless known as a facility open toindustrial research. To accompany thedevelopment of the ESRF industrialpolicy, it is necessary to give theindustrialists more information aboutthe scientific potential of synchrotronradiation for industrial applications.

A good opportunity to accomplishthis was given by the exhibition«Physique 97» which was held in Parisfrom 7 to 9 October 1997. The ESRFdecided to participate in this exhibitionand the operation was managed bythe Industrial Relations Office, withsupport from the Information Office.

An exhibition such as «Physique97» attracts a large number ofscientists as well as visitors fromindustry (the ESRF stand receivedmore than 1000 visitors over the threedays, 50% from industry, 50% fromresearch institutes).

In addition, this exhibition beingprincipally commercial (most of theexhibitors are companies proposingtechnological and scientific equipment),it is a perfect place to find partners forthe marketing of instruments developedat the ESRF.The ESRF stand consisted of:• scale models to show and explain thefunctioning of the machine and ofbeamlines;• posters (4 general posters about theESRF, 3 posters about technical

developments and 5 posters aboutrecent scientific results);• some instruments developed atthe ESRF (the FRELON camera,monochromator crystals and thecryogenic cooling system formonochromators).

The stand thus combinedinformation provided by the posters withmore animated areas: the interactivescale models, the demonstration ofthe FRELON camera in real operation

and of course lively discussions withthe ESRF staff present on the stand.General and technical documents werealso distributed in large quantitiesduring the three days.

The ESRF stand was particularlyappreciated by the visitors who couldfind all kinds of information. It wasalso strongly congratulated by theorganizers of the exhibition (SociétéFrançaise de Physique).

J. Doucet

ESRF AT «PHYSIQUE 97» IN PARIS

The Delegate of the French Minister of Public Education,Research and Technology, M. Descusse (right) shows a strong interest

in the remarkable scientific results and developmentsexplained by J. Doucet (second from right).

WORKSHOP

«CHEMISTRY AND SURFACES» FOR INDUSTRIALISTS

4 June 1998, ESRF, Grenoble

In the series of workshops «ESRFfor Industry», which are organizedfor industrialists from Europeancompanies, a workshop dedicated to«Chemistry and Surfaces» will takeplace at the ESRF on 4 June 1998. Forthe first time, it will be held togetherwith a similar and complementaryworkshop «ILL for Industry», on 5June 1998.

The morning program willcomprise a general presentation of theESRF, a description of the variousmodes of access to the facilities andpresentations of the possibilities ofcharacterization of materials andcontrol. The topics are: diffractionanalysis of powdered materials;chemical reactions studied byabsorption techniques; surface,

interface and thin-layer analysisusing synchrotron radiation; as wellas new opportunities in micro-imaging techniques (tomography, x-raymicroscopy, phase-contrast technique).In the afternoon, a visit of thebeamlines will be organized duringwhich the participants will benefitfrom informal discussions withscientists from the ESRF.

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

10

27TH AND 28TH COUNCIL MEETINGS

At its 26th meeting on 10 and 11June 1997, the Council electedG. von Klitzing as its Chairman forthe period from 1 January 1998 to31 December 1999. He succeedsProf. F. Menzinger who has chairedthe ESRF Council since January1996.

G. von Klitzing, 64, has studiedlaw and economics. His first contactwith research and scienceadministration was in 1962 whenhe entered the former GermanFederal Ministry of Nuclear Energy(from which arose, after severalintermediate steps, the BMBF oftoday) where he was involved in

European Space Research.G. von Klitzing’s career after this

was marked by a succession of periodsat European organizations such asCERN, ESRO and EURATOM, andtasks within the Federal Ministryof Research and Technology.During these latter periods he wasa member of various committeesat European level (EuropeanCommunities, JET Council, WorkingGroup «International Fuel CycleEvaluation», ...).

From the end of 1991 to mid-1997, G. von Klitzing was DeputyChairman of the Board of Directorsof the Forschungszentrum Jülich, i.e.

the German Member of the ESRF.During this period he became

involved in ESRF matters, first as amember of the German delegationto the Council, from mid-1993to mid-1995 as Chairman ofthe Administrative and FinanceCommittee and, for the last twoyears, as Vice-Chairman of the ESRFCouncil.

The new Vice-Chairman of theCouncil is Dr Paul E. Zinsli from theSwiss Federal Office of Educationand Science, who - with more thanten years of service - has been thesenior Head of Delegation at theESRF.

The ESRF Council met on 17October 1997 in Paris in order tofinalize the wording of theArrangement between the ESRFand the Government of Portugal onthe long-term scientific use ofsynchrotron radiation (see alsoESRF Newsletter N° 29, p. 3). Theapproved Arrangement was signedon 12 November 1997 in Lisbon andbecame effective on 1 January 1998.

At its meeting on 4 and 5December 1997 in Grenoble, theCouncil approved the Medium-TermScientific Program submitted byManagement which, apart from thecontinuous efforts to improve theradiation source and the variousbeamline components (e.g. optics,sample environment, data acquisition),comprised some specific projectsaiming at an extension of experimentaloptions at the ESRF, such as a TRXFstation for silicon wafer mapping, therebuilding of the MAD beamline on aninsertion device location and theestablishment of a second nuclearresonance scattering station.

The Council discussed its position

vis-à-vis further CollaboratingResearch Group beamlines (regardingtheir scientific objectives and theburden on the ESRF generalinfrastructure). It decided that for anyfurther CRG beamline, the CRG incharge will have to bear the fullinvestment cost of the correspondingbeamline front-end, and set 16 CRGbeamlines branches as the thresholdfor the next review of the situation.(Currently 13 branches are inoperation or under construction ascompared with 34 branches of ESRFpublic beamlines.)

The Council noted some plannedbuilding construction measures, themost important being the extension ofthe Guest House by a third wing,possibly with the participation of theInstitute Laue Langevin.

The budget for 1998, unanimouslyapproved by the Council, provides foran expenditure level of 417 420 kFF,covered by Members’ contributions of396 000 kFF and 21 420 kFF of otherincome.

The Council took note of theMedium-Term Financial Estimates

presented by Management andconsidered the expenditure and incomeprofiles suggested for the period up to2002 as a reasonable guideline forfurther developing these estimates andthe associated scientific program.

With a view to the end of the ESRFconstruction period, the Councilapproved a revised version of theFinancial Rules. It also adopted a firstpart of «Guidelines for a Readjustmentof Contribution Rates», essentiallydealing with the assessment ofthe scientific use which is made of theESRF by the scientific communitiesof the various Contracting Parties. (Inthe case of a lasting and significantimbalance between use andcontributions, the ESRF Conventionprovides for the option of areadjustment of the contribution rates).

Finally, the Council slightlyamended the Procedures for beamtime allocation in order to enableManagement to better accommodateprojects involving industry-basedproposers and providing potential forindustrial applications.

K. Witte

GEORG VON KLITZING

NEW CHAIRMAN OF THE ESRF COUNCIL

OPEN DAYS

NEWSLETTER IN BRIEF

ESRF - APRIL 1998

11

After nearly four years without anOpen Day, the ESRF opened its doorsagain and attracted about 2000 peopleon 14 and 15 March 1998.

On their tour through the ESRF, the

visitors had the opportunity to discover12 «stations», where ESRF staff wereposted and gave explanations: Therewas a general presentation, and theycould visit the machine (the roof of the

storage ring had been partly removed)and several beamlines (surfaces,materials, biology and medicine).The visitors could also attend ademonstration, on a scale model, of anew experimental station for industrialuse in micro-electronics. The visitorsstayed 3 hours on average and seemeddelighted with their visit.

THEATRICAL ITINERARY «LUMIÈRES»

VACANCIES AT THE ESRF ON 20 MARCH 1998

Ref Subject

SCIENTIST 2170 High brilliancePOST-DOC PDID15B M5

PDID9 DiffractionPDID21 X-ray microscopyPDID10A Troika group

ENGINEERS 6106 Buildings/infrastructure2206 Software supportCDD/CH Digital electronics

TECHNICIAN 2521 Nuclear resonance

OTHER 6103 Survey and alignment

If you are interested,please send us a fax(+33 (0) 4 76 88 24 60)or an e-mail(brink@esrf.fr)with your address,and we will provideyou with anapplication form.You can also printout an applicationform on theWorld Wide Webhttp://www.esrf.fr

Previous postdoctoralexperience withsynchrotron radiationis essential.

On the initiative of the director ofa theatrical company in Grenoble, sixactors and one musician wereinvolved in a «theatrical itinerary»through the ESRF, mixing art andscience. From 20 to 31 March 1998,twice a day, this theatrical itinerarywas proposed to a group of about 35people, who could hear about scienceand synchrotron light, as well as

poetic, philosophical and literary texts.This project was strongly

supported by institutions such as theRhône-Alpes region, the Grenobletown council and the Isère departmentcouncil.

It had an excellent impact in themedia, both in the press (30announcements and articles) and onradio/TV (about 25 broadcasts).

Foreign newspapers, radio and TValso showed a great interest in thisevent. On Thursday 19 March, forthe last rehearsal, we welcomedcorrespondants from Ny Teknik(Sweden), La Tribune de Genève(Switzerland), TV3 Barcelona and ElCorreo (Spain), Giornale di Brescia(Italy), NRC Handelblad (TheNetherlands) and RTBF (Belgium).

Looking into the storage ring.

Following explainations on a scale model.

MAGNETISM WORKSHOP

ESRF - APRIL 1998

13

X-RAY SCATTERING AND MAGNETISMC. VETTIER1 AND G. H. LANDER2

1 - ESRF, EXPERIMENTS DIVISION

2 - KARLSRUHE, GERMANY

agnetism has become a major field ofinvestigation with synchrotron radiation. Inparticular, the highlights of the ESRF for1996/1997 state that «The use of x-rays for theinvestigation of magnetic properties of matter isa relatively recent field», and then continueswith 11 pages of various examples of neweffects measured at the ESRF in this field. A largefraction of the allocated beam time is spent onstudying various aspects of magnetism. To focuson this activity a workshop was organized at theESRF and held over the weekend of 15/16November 1997 before the large SR50conference. Some 90 attendees came to hearspeakers over the day and a half of meeting.M. Blume (BNL, USA) opened the conferencerecalling the fundamental cross-sections andshowing how the general non-resonance cross-section was modified by the introduction ofresonant terms. B. Gyorffy (Bristol, UK) andM. Brooks (Karlsruhe, Germany) discussed newdirections in theory stimulated by synchrotronexperiments. We heard reports of experimentsusing both non-resonant (Th. Brueckel, Jülich,Germany; J. McCarthy, ESRF) and resonanttechniques (C. Detlefs, Ames, USA;S. Langridge, Rutherford Lab, UK) that gaverepresentative examples in the field of scatteringfrom synchrotrons in Hamburg, Brookhaven andGrenoble, and included new experiments tolook for «orbital ordering» in perovskitematerials (S. Ishirara, Sendai, Japan). Later thefirst day we turned to dichroism with a theorytalk by P. Carra (ESRF) on the sum rules, andcontributions by G. van der Laan (DaresburyLab, UK), A. Rogalev (ESRF), and the extensionof dichroism to photoemission (G. Kaindl,Berlin) and magnetic EXAFS (G. Schütz,Wurzburg, Germany). These talks showedthat techniques using the dichroic signal arerapidly expanding and that more advancescan be anticipated.After a long first day, the participants wererewarded with a fine dinner in town, and the

second day started in earnest at 9am (onSunday) with contributions on surfaces andfilms. N. Bernhoeft (ILL), D. Gibbs (BNL, USA),G. Helgessen (Kjeller, Norway) and C. Sutter(ESRF) showed the variety of that are now beingexamined with scattering, almost all usingresonant techniques and involving rare-earthsand actinides. The workshop ended withreviews of inelastic scattering (F. Sette, ESRF)and perspectives for the future (M. Altarelli,ESRF). These last talks introduced new areas andshowed that, despite the considerable impactof synchrotron studies on magnetism, the fieldis rapidly expanding and the future looks brightand challenging.All participants could enjoy a fruitful exchangeof ideas and experiences. The interest andinvolvement of the audience was reflectedin lively discussions. Although there wereno neutron talks per se in the workshop, allattendees were aware of the significant impactneutrons have had, and continue to have,on magnetism. It was appropriate that muchof the complementary nature of the photonprobe should be emphasized in the ChadwickAmphitheatre of the Institut Laue Langevin –under the stern gaze of the discoverer of theneutron. It is planned to organize a commonworkshop in the near future together withthe neutron community on the complementaryuse of neutrons and x-rays in the study ofmagnetism.We illustrate the broad scientific range of theworkshop with the following two articles thatexpose recent advances in the field, althoughthey are not directly related to magnetism.In a first paper M. Altarelli describes the physicsof the ordering of electronic orbital occupancy,which shows some similarities with magneticorder, and N. Bernhoeft et al. present thecompetition between the coherence lengthof the probe and correlation length scales inthe sample that lead to new effects in x-rayand neutron scattering experiments.

M

MAGNETISM WORKSHOP

ESRF - APRIL 1998

14

DETECTION OF ORBITALORDER IN TRANSITION

METAL OXIDES BYRESONANT X-RAY

SCATTERING

M. ALTARELLI

ESRF, EXPERIMENTS DIVISION

The transition metal oxides havefascinated chemists and physicists formany decades with their intriguingstructural, magnetic and electronicproperties. One of the remarkablephenomena displayed by a number ofthese compounds is the Jahn-Tellereffect, in which an orbital degeneracy ofthe transition metal ion is lifted by alow-symmetry lattice distortion, which«chooses» one of the degenerate states.In a lattice of such ions one canencounter a cooperative

Jahn-Teller effect, in which theoccupation of the degenerate orbitalsdisplays long-range order, characterizedby a well-defined wavevector. Forexample in LaMnO3, the Mn ionsare partitioned in two sublattices, inwhich the 3d(3x2-r2) or, respectively, the3d(3y2-r2) orbitals are occupied. It turnsout that orbital order can be favourednot only by the consideration oflattice distortions, but also by othermechanisms, notably by magneticinteractions. This is because the relativeorientation of occupied d orbitals inneighbouring ions determines thestrength and sign of the indirectexchange (superexchange) interactionbetween their spins.

Recently, in view of the high currentinterest on transition metal oxidesstimulated by the discovery of cupratehigh-temperature superconductorsand of colossal magneto-resistancein the manganates, orbital orderphenomena have been paid muchattention. One aspect which wasdiscussed during the ESRF workshopon Magnetic Scattering is thepossibility of a direct observation oforbital order by resonant x-rayscattering at the K edge of the transitionmetal. This could give a direct handleto the order parameter of this peculiarkind of ordered state, and a measure ofits temperature dependence, and therelationship of its onset to structural ormagnetic phase transitions, thuscontributing to an understanding of itsorigin. In fact there is up to date no

satisfactory way to observe andmeasure the orbital order parameter,although in principle non-resonantdiffraction should detect thecorresponding change in the shape ofthe electron density. The correspondingsignal is however expected to be weak(less then 0.1 electrons per unit cell).

In some recent theoretical workdone at the ESRF by M. Fabrizio,M. Altarelli and M. Benfatto, the idea isto exploit the sensitivity of resonantscattering of polarized x-rays to theavailability of empty intermediatestates of the appropriate symmetry.In an orbitally-ordered system, theyare available on one sublattice but noton the other. In this work theparticularly intriguing case of V2O3was considered; in this compound, thewavevector of the orbital order hasnever been experimentally measuredand is not unambiguously determinedby theory; recent indirect evidence, isin some cases, interpreted as stronglysupporting the existence of orbitalorder, and in some others as putting thevery existence of orbital order inquestion. The conclusion of theAuthors, which has not yet beenverified, is that orbital order should bedirectly observable by resonantscattering and therefore all controversycould be settled by experiments.

In experiments performed byY. Murakami et al. on La0.5Sr1.5MnO4as well as on LaMnO3, a signal atthe wavevector of the expectedorbital order was detected, which wasstrongly resonant at the dipole edgecorresponding to the Mn 1s to 4ptransitions. In theoretical work done inSendai by S. Ishihara and S. Maekawa,the sensitivity to orbital order wasascribed to a modulation of the energyof the empty 4p levels depending onthe selective occupation of the3d orbitals. These results of theJapanese groups were reproducedin Brookhaven by D. Gibbs andM. Blume. In our opinion, more workis called for, in order to verify whetherthe 4p level energy modulation iscontrolled by other factors or not, suchas the displacement of the ligandoxygens.

In summary, the workshop provideda very timely opportunity to discuss thecontribution of x-ray scattering to theinvestigation of orbital order, bringingtogether the groups involved in thisproblem at the ESRF, at NSLS and inJapan.

PROBE COHERENCEVOLUMES AND THE

INTERPRETATION OF SCATTERINGEXPERIMENTS

N. BERNHOEFT2, A. HIESS2,A. STUNAULT1, S. LANGRIDGE5,

D. WERMEILLE1, C. VETTIER1,G. H. LANDER3, M. HUTH4,

M. JORDAN4 AND H. ADRIAN4.

1 ESRF, EXPERIMENTS DIVISION

2 ILL, GRENOBLE

3 EITU, KARLSRUHE

4 UNIV. OF MAINZ

5 RAL, UK

One of the significant features ofsynchrotron radiation, especially thatfrom 3rd generation insertion devices, isthe small-energy bandpass of theemitted radiation. In the case ofundulator beams monochromatized bySi crystals at the ESRF, the wavelengthbandpass (∆λ/λ) attains ~10-4. As aresult the photon beam exhibits partialcoherence with a longitudinal coherencelength ξL = λ/(∆λ/λ) of the order of1 µm for 1 Å x-rays. The possibility ofcoherence over such dimension isfundamental to the experimentsinvolving «speckle» patterns at theESRF, but it is also an importantconsideration when conventionalscattering experiments are performed,especially when the quantity ξLbecomes comparable to the dimensionsof the material being studied, or with theextinction length; the latter being writtenas 1/µ and corresponding to the lengthover which absorption reduces theintensity of the radiation by a factor 1/e.Within a coherent volume theamplitudes rather than the intensities ofthe scattering process must be added.We emphasize here that these ideas areindependent of the perfection of thecrystal scattering the photon beam, i.e.they apply equally to both the kinematicand dynamic theories of diffraction,although the treatment in the two casesis, of course, different.

The standard procedure forcorrecting for absorption involvescorrecting the intensities diffractedfrom a volume of scattering, i.e. theinterference between the amplitudes isfirst calculated to give the beamamplitude for a given reflection, andthen the absorption correction is madeon the final intensities. This isappropriate when the extinction length

MAGNETISM WORKSHOP

ESRF - APRIL 1998

15

is large and the coherence volume issmall, 1/µ > ξL, such as when alaboratory source is used and thewavelength is far from an absorptionedge. However, it is inappropriate whenresonant scattering experiments (i.e. thephoton energy is tuned to an absorptionedge) are performed at a synchrotronsource. For example, at the actinide Medges µ ~ 20,000 cm-1 so that 1/µ ~ 0.5µm and clearly 1/µ < ξL in this case.

These ideas have been pursued andnumerical routines developed tosimulate possible situations. Space doesnot permit a complete description of thetheory, but we will illustrate how theseideas influence scattering patternsactually observed on the magneticdiffraction beamline, ID20.Magnetic resonant scatteringoccurs when the photon beam istuned to a particular resonanceand a large enhancement of the(usually weak) magnetic cross-section occurs. The largest sucheffects have been found to beassociated with actinide Medges, when the enhancement ofthe magnetic cross-section canreach a factor of ~ 106. At thesame time the absorption is suchthat the probe becomesessentially one of the nearsurface region. In this extremesituation, numerical simulationsshow that the exact form of thescattering, especially its energydependence, which changes theextinction length, will depend onthe spatial localisation of the

scattering centres. This is a newconcept, coupling the spatial aspectwithin the sample being examined tothe energy dependence of the scatteringintensity.

The first example is taken fromexperiments on thin films of theantiferromagnet UPd2Al3. The filmshave thicknesses (t) 800 and 1600 Å andwere grown epitaxially on the (111)surface of LaAlO3 by co-evaporation ofthe elements. The antiferromagnetic(AF) order of the compound isalternating (001) layers of ferromagneticspins so that the new Bragg peaksproduced by the AF repeat distance are(0, 0,n/2)where n = odd integer. Figure 1shows the results of measuring the

energy-dependence of the scatteredintensities for two different values of n,and for the two films. There is a cleartendency for the narrowest energywidths to be achieved when the photonbeam has the shortest optical path in thematerial. The dashed curve is the resultof numerical simulations consideringthat the probe and sample transversecoherence may be neglected; as shown,the simulations capture the essentialbehaviour. At T = 4 K these films areordered throughout their thickness, andprovide a good test of the theory becauseof their limited thickness. Just below theordering temperature TN only the toppart of the films orders magnetically andthis may be seen by a 20% narrowing ofthe energy width of the (0, 0, 1/2)reflection for the 1600 Å film. The bestfit to this energy profile suggeststhat only the top 600 (± 50) Å of thefilm is magnetically ordered at thistemperature. The lack of homogeneity isassumed to arise from the 0.15%interfacial strain between the film andthe substrate.

The second example is taken fromexperiments on another uraniumantiferromagnet UAs. The experimentsare performed on an infinitely (severalmm) thick crystal. In Figure 2, we showthe energy profile as a function ofphoton energy when the crystal is nearTN. The profile clearly shows splitpeaks. This cannot arise from orderingjust at the surface, as was seen above inthe films, and which always gives riseto a narrowing of the energy width.Instead in this case there is a non-diffracting layer of some 400 Å at the

6.0

5.5

5.0

4.5

4.0

3.5

3.0

2.5

2.0

t/sin(θ)[Å]

HW

HM

[eV

]

UPd2Al3 films

102 103 104 105

UPd2Al3 films

uranium M4 resonance

bulk value

1600 Å(001/2)800 Å

(001/2)

1600 Å(003/2)

800 Å(003/2)

Fig.1: Energyhalf-width of theantiferromagneticpeaks (all atT = 4 K) at the M4absorption edge asa function of opticalpath length (t = filmthickness, and θ isthe Bragg angle fora reflection) of thephoton beam in thethin films ofUPd2Al3.The dashed curveis the numericalsimulation assumingthat the orderingis establishedthroughout the filmthickness.

0.10

0.08

0.06

0.04

0.02

E [keV]

Cou

nts

/ min

3.70

UAs (102) 100 KM4 resonance

3.71 3.72 3.73 3.74 3.75

Fig. 2: Energy profile of an

off-specularantiferromagnetic

reflection from UAstaken near TN on

ID20. The solid lineis a fit to the data

employing the modeldescribed in the text

and assuming thatthere exists a non-diffracting layer of400 Å of the same

material on top of thecrystal. This belongs

to another domainand so diffracts into a

different direction.

MAGNETISM WORKSHOP

ESRF - APRIL 1998

16

surface of the UAs crystal that actuallycorresponds to another magnetically-ordered domain not probed by the off-specular (102) reflection. The solid lineis a fit, and it reproduces the data verywell.

In conclusion, we have introducedthe idea that the «probe coherencevolume» plays an important part inthe interpretation of diffraction datataken at the ESRF. This representsa re-examination of assumptionsconventionally used to interpretdiffraction data. In most cases the

probe coherence volumes are small –special efforts have to be made toincrease the monochromaticity of thebeam, which usually result in anunacceptable loss of intensity.However, these new conditions areencountered with synchrotron sources,especially with the new undulatorbeams, and care must be taken tounderstand the consequences. In thepresence of strong absorption, neweffects arise and we have shown thatin certain cases the energy dependenceof the intensity may be used to identify

the spatial localization of thescattering centres. This opens the wayfor further experiments exploiting thepartial coherence of the incident x-raybeam. The coherence of x-ray sourcesand the narrow bandpass beams thatare available from undulators at 3rdgeneration sources make these effectsmore qualitatively important thanpreviously thought. Discussionsduring the workshop made this pointclear and further emphasised the newinformation that can be obtained fromexperimental results.

THE 17TH GENERAL CONFERENCE OFTHE CONDENSED MATTER DIVISION

OF THE EUROPEAN PHYSICAL SOCIETY

will be held as a joint meeting with the

6e «JOURNÉES DE LA MATIÈRE CONDENSÉE»OF THE FRENCH PHYSICAL SOCIETY

at Grenoble (Univ. Campus)from 25 to 29 August 1998

e-mail (for information): cmd17@labs.polycnrs-gre.fr • • • e-mail (for abstracts): abs17@labs.polycnrs-gre.fr

CONFERENCE SECRETARIAT

Laboratoire Louis Néel, BP 166 X F-38042 Grenoble cedex 9 (France)Phone: 33 (4) 76 88 90 02 - Fax: 33 (4) 76 88 90 04

It will comprise seven plenarylectures given by the followinginvited speakers: G. Abstreiter (Si/GeNanostructures), M. Bruel (Smart Cutof Si), O. Fischer (Superconductivity),J. Pendry (Quantum Friction),Y. Petroff (Synchrotron Radiation),J. Prost (Biophysics) and R. Scherm(Neutron Scattering).

The last plenary session will bededicated to the Hewlett-Packardprize-giving.

Semi-plenary sessions (50 invitedspeakers) will also be organised aswell as 37 parallel mini-colloquia,selected after a survey among thephysicists.

Junior scientists (PhD students) arewarmly encouraged to participate.Grants will be offered to cover most ofthe accommodation expenses andregistration fees.

An exhibition of books andscientific equipment will be heldduring the conference.

We will particularly welcomeany initiative in favour of theemployment of students andexchange opportunitites for postdocs.A round-table discussion on jobopportunities after PhD studies willbe organized.

Finally, a word about Grenoble.Besides its outstanding scientific

facilities, it has plenty of attractiveaspects you may wish to discover.It contains several major museumsand the center of the city is very livelyat night. At the end of summer,hiking is very appealing and can bedone in an area spread over threemain mountain ranges which are allwithin half-an-hour by car of the citycenter.

The program of the conference ison the Web at http://www.polycnrs-gre.fr/eps.html

EARLY REGISTRATION15 MAY

X-RAY IMAGING WORKSHOP

ESRF - APRIL 1998

17

MEETINGS ON X-RAY IMAGING

-ray imaging wasone of the maintopics of the seriesof meetings held inNovember 97 inGrenoble.They started with theSR-50 celebration ofthe 50th anniversaryof synchrotronradiation, where thelast day was nearlycompletely devotedto imaging.The User's Meetingfollowed, one dayafter; it included aplenary talk on thein-situ phasetomographicinvestigation of damage in a composite alloy,(J.Y. Buffiere), and a «mini-workshop» onimaging. The satellite meeting on x-rayimaging held the next day at the ESRFconcluded these scientific events. The 'imaging' mini-workshop, organized byJ. Baruchel within the frame of the User'sMeeting, gave a general view on what hasbeen achieved at the ESRF (M. Schlenker'sreview paper), and on the big variety oftechniques which have been developed at thevarious beamlines partly or totally devoted toimaging (BM5, A. Souvorov; ID17, P. Spanne;ID19, J. Härtwig; ID21, J. Susini; ID22, C. Raven).The use of phase contrast was highlighted byseveral speakers, and the specificities of thebeamlines emphasized. A presentation of theavailable detectors (J.P. Moy) and two shortscientific highlights, on the joint use of phaseradiography and diffraction imaging to investigatedefects in quasicrystals (L. Mancini), and on thepeculiar deformations of the ice of the Antartica(P. Duval) concluded this mini-workshop.The satellite meeting held at the ESRF wasoriginally intended to bring together the varioususers of the 'long' ID19 beamline. ID19 was thefirst operational beamline completely devoted tox-ray imaging. The idea was, 16 months afterthe official aperture to users, to show what waspossible, which topics were already investigated

and what are the main results. The success ofthe meeting went beyond what was originallyexpected: more than 80 scientists attendedthe sessions and more than 50 posters weredisplayed. These posters presented an importantfraction of the work performed on ID19, butalso imaging work performed on other ESRFbeamlines (BM5, ID22) and even othersynchrotron radiation facilities (Daresbury,Hasylab). This one-day meeting started witha presentation on the new possibilities alreadyimplemented. It was followed by a series ofshort oral presentations of the posters. The mainof the afternoon was devoted to an animatedposter session, and a visit, for the interestedpeople, of the ID19 and ID22 beamlines.The highlights and trends of this meetingwere extracted by B. Tanner. His, of coursesubjective, but quite largely shared, highlightslist included the first microfluorescencemapping, the microtomographic work (bones inabsorption, Al-SiC composites and insects usingphase radiography), the combination of phaseand diffraction imaging (defects in quasicrystals,domains in lithium niobate), the applications ofdiffraction imaging to magnetic materials and,through Moire imaging, to semiconductors.The three short papers which follow partlyillustrate these highlights.

X

J. Baruchel

X-RAY IMAGING WORKSHOP

ESRF - APRIL 1998

18

COMBININGDIFFRACTION

TOPOGRAPHYAND PHASE IMAGING

J. BARUCHEL, P. CLOETENS*,M. DRAKOPOULOS, L. MANCINI ,

P. REJMÁNKOVÁ-PERNOT,A. SNIGIREV,

I. SNIGIREVA, A. SOUVOROV

ESRF, EXPERIMENTS DIVISION

* ALSO AT EMAT, RUCA, ANTWERP, BELGIUM

This work results from collaborationbetween ESRF staff members and

scientists from other institutes: J. Gastaldi and E. Reinier (CRMC2,

Marseille), J.P. Guigay andM. Schlenker (Lab. L. Néel, CNRS,

Grenoble), Z. Hu and P. Thomas (Univ. of Warwick),

P. Moretti (Univ. C. Bernard, Lyon).

The deliberate combination of phaseand Bragg-diffraction imaging hasrecently led to important results, givinga new impulse to the use of thesetechniques. The experiments are beingcarried out on several beamlines: BM5,ID11, ID19 and ID22.

In the classical version of x-raydiffraction imaging ('topography') thesingle-crystal specimen is set to giveBragg reflection(s), and adetector placed as nearas possible records thediffracted beam(s). Contrastshows up whenever thecrystal includes featuresthat locally affect theBragg reflectivity. Thuscrystal defects (dislocations,inclusions, ferroelasticdomains,...) are made visibleand can be characterized.The investigation intransmission of thickcrystals often requiresrestricting one dimensionof the beam («section»topography): the imagethen originates, in a firstapproximation, from avirtual slice of the crystal.When working at theESRF an additionalparameter can be varied toobtain further information:the 'propagation' distancebetween sample anddetector.

The 'propagation' parameter alsoallows phase radiography andtomography to be performed in a verysimple way. The 'propagation' techniquetakes advantage of the small size of thex-ray beam sources at ESRF, yielding avery small divergence α of the beam asseen from any point in the specimen.This results in a high degree of spatialcoherence, characterised by thetransverse coherence length lc = λ/2α,where λ is the x-ray wavelength.Local variations in the x-ray opticalpath-length in the sample, i.e. phasevariations across the beam, thenproduce contrast through Fresneldiffraction. In the case of negligibleabsorption the image is uniform whenthe detector (x-ray film in our case) isclose to the sample. As the detector ismoved further away, contrast builds up,and features on different length scalesare brought out more conspicuouslydepending on the specimen-distance D.The edges of an inhomogeneity of size sare outlined when D < s2/λ, whereasinterference fringes occur, the imagebeing a hologram, for further increaseof D.

The investigation of defects in Al-Pd-Mn single quasi-crystals is a firstcase where the combination oftopography and phase radiographyallowed a breakthrough. It is nowfamiliar that quasicrystals, non-periodic

solids with long-range order, exist, andproduce fine diffraction peaks for x-raysjust as do crystals. Therefore Bragg-diffraction imaging is applicable tosingle quasicrystals too, and it hasalready produced some informationabout the structural defects in thesematerials, with topology quite differentfrom that of crystals. Figures 1a and bshow a region of a single quasicrystalslab, imaged in simple transmission,with no strong Bragg reflection excited.In the enlargement (Figure 1a) a cleardelineation of internal holes with ashape featuring, just as etch pits incrystals, the icosahedral point symmetryof the material, and elongatedprecipitates, are observed. Figure 1cshows a topograph of the same region asin Figure 1b. Loop shaped defectimages, corresponding to distortedareas, are visible. The geometricalrelations between the two, apparentlyquite different, images of the sameobject can be determined by recording aphase radiograph with the sample set fora strong Bragg reflection (Figure 1d).Joint use of these images allows todetermine the relationships between theholes or precipitates and the complicateddefects around them [1].

In conventional Bragg-diffractionimaging, the surfaces of the samples(and/or monochromator) should be strainfree, but no specific demand is made on

Fig. 1: Imagesof an Al-Pd-Mn

quasicrystalrecorded at

λ = 0.35 Å a)phase radiograph

(D = 0.5 m).In b), c) an d)

the same regionof the sample ispresented (D =0.2 m) b) phase

radiograph withno strong Bragg

diffraction excite;c) topograph

(diffracted beam)and d) combined

phase radiographand topograph

(transmittedbeam); sample

set for the τ3

(0/0 0/-2 0/0)reflection.

a b

c d

X-RAY IMAGING WORKSHOP

ESRF - APRIL 1998

19

their flatness. This is no more true whendealing with high quality crystals andusing a coherent x-ray beam. Thediffraction image then carries phaseinformation, yielding extra contrastwhen the optical path length is irregular.Faint scratches on the surface of silicon

crystals used as monochromators wereshown to produce strong phase imageswhich are not visible in 'classical'topography. The same approach wasused to reveal thin oxide strips onreflections where the associateddistortion does not produce contrast [2].

The combination of phase anddiffraction imaging led to spectacularresults with beams Bragg-diffracted byperiodically poled lithium niobatecrystals. These crystals exhibit largeoptical nonlinearities and are widelyused for electro-optic and acousto-opticapplications as well as for opticalfrequency conversion. Poling producesa periodic spatial distribution of the signof the nonlinear optical coefficient,crucial for the quasi-phase matchingrequired for optical harmonicgeneration. When parts of the coherentx-ray wave front diffracted fromdomain inverted regions are split andsubsequently overlap, they interfere [3].Figure 2 shows a series of images atdifferent sample-to-film distances, inBragg (reflection) geometry. Thefringes show the phase origin of thephenomenon. The inverted ferroelectricdomains exhibit structure factors withvery similar modules, but differentphases. This gives rise to contrastmechanism, used to image the domaindistribution in the bulk in Laue(transmission) geometry as shown onFigure 3. Many reflections involvenearly no effect of domain-relatedlattice distortion. Contrast then mostlyoriginates from the phase shift betweenthe structure factors in the domains. It isthus possible to extract, from imagesrecorded at different distances, a directmeasurement of this phase difference[4]. It was on the other hand shown thatthe contrast of the domain images canbe enhanced by the application of anelectric field [5].

The above examples show some ofthe unique possibilities resulting fromthe combined use of Bragg and Fresneldiffraction imaging techniques. This isclearly an emerging new field.

REFERENCES

[1] L. Mancini, E. Reinier, P. Cloetens,J. Gastaldi, J. Härtwig, M. Schlenker,J. Baruchel, Phil. Mag., in press.[2] a) A. Souvorov, I. Snigireva, A. Snigirev in«Highlights in X-Ray Synchrotron RadiationResearch», pages 8-9 and 101, Grenoble, 1997b) A. Snigirev, M. Drakopoulos, S. Kuznetsov,I. Snigireva, A. Souvorov, submitted.[3] Z. Hu , P. Thomas, A. Snigirev, I. Snigireva,A. Souvorov, P. Smith, S. Teat, Nature, in press[4] P. Rejmánková-Pernot, P. Cloetens,J. Baruchel, Jean-Pierre Guigay, P. Moretti,submitted.[5] P. Rejmánková, J. Baruchel, P. Moretti,M. Arbore, M. Fejer and G. Foulon, J. Appl.Cryst, in press.

Fig. 2: Images of inversiondomain walls in aLiNbO3 crystalrecorded using the006 reflection(reflectiongeometry), λ = 1 Å,at different sample-to-film distancesD: a) 0.5 cm b) 20 cm c) 50 cm.Note theinterference fringeswhich build up asthe distance isincreased.

Fig. 3: Section topographs (transmission geometry) of a periodically poledLiNbO3 crystal using the 274 reflection, λ = 0.24 Å, recorded at differentsample-to-film distances D: a) 0.11 m, b) 0.51 m, c) 1.67 m, g is theprojection of the diffraction vector on the detector.

a

b

c

a

b

c

X-RAY IMAGING WORKSHOP

ESRF - APRIL 1998

20

IN SITU DAMAGEASSESSMENT IN MICRO-

HETEROGENEOUSMATERIALS USING HIGH

RESOLUTION X-RAYTOMOGRAPHY

J-Y. BUFFIERE1, E. MAIRE1,P. CLOETENS2*, G. PEIX3, M. SALOMÉ4,

J. BARUCHEL2.1 GEMPPM UMR CNRS, VILLEURBANNE, FRANCE

2 ESRF, EXPERIMENTS DIVISION

3 CNDRI INSA, VILLEURBANNE, FRANCE

4 CREATIS INSA, VILLEURBANNE, FRANCE

* ALSO AT EMAT, RUCA, ANTWERP, BELGIUM.

Metal Matrix Composites (a metalreinforced by ceramic particles orfibers) combine good mechanicalproperties and a low density. Hence,those materials are very attractive forautomotive applications where a gain inweight is highly desirable. However, thepoor fracture properties of thosematerials have greatly restricted, so far,their use in industrial applications.Therefore, in the last years, muchattention has been paid to the study offracture processes of Metal MatrixComposites (MMC). For instance, insitu mechanical tests have been used byseveral authors and provide veryinteresting information on the loss ofductility of MMC. However, therelevance of surface examination has tobe checked by destructive observationswithin the material through timeconsuming techniques like serialsectioning which are not free ofartefacts. To overcome this problem, wehave used phase contrast tomography, toobtain three dimensional images ofdamage within materials under stress.

The damage mechanisms of anAluminium/Silicon Carbide (Al/SiC)composite strained in tension have beenstudied in situ by phase contrasttomography using a specially designedtensile testing device. The phase contrasttechnique enhances the contrast betweenthe aluminium matrix and the SiCparticles with respect to classicalattenuation-based tomography. Thesteady development of cracks in the SiCparticles has been monitored on the samesample for increasing values of plasticdeformation. The use of phase contrastenables to visualize cracks with anopening lower than 0.5 microns in theSiC reinforcements. For the first time, aquantitative comparison of the damage

mechanisms observed at the surface andthose observed in the bulk, by a nondestructive method, is presented.

EXPERIMENTS

High-resolution x-ray tomographyexperiments were carried out at the ESRFin Grenoble on the «topography»beamline ID19. Images of damageinduced by a tensile test within a MMC(aluminium matrix reinforced by siliconcarbide (SiC) particles) have beenobtained.

Because of the low difference in thex-ray attenuation of aluminium and SiC,classical transmission tomography,based on attenuation laws, was hardlyable to discriminate between matrix andreinforcement. To improve this contrast,the very high lateral coherence of thebeamline ID19 was used to producephase contrast images through a verysimple experimental set-up [1]. A FastRead-Out Low Noise CCD detector,developed at the ESRF, was placed atabout 1 metre behind the sample torecord edge diffraction patterns resultingfrom discontinuities in the material suchas interfaces between reinforcementand matrix or strain-induced cracks.Figure 1 shows a three-dimensionalreconstruction obtained from therecorded phase images, where thereinforcing particles are clearly visible.

The studied material was a 6061aluminium alloy reinforced with 10 %in volume of silicon carbide (SiC)particles with an average size of 120µm. The material was producedthrough a rheocasting route andsubsequently extruded at hightemperature. A special tensile testingdevice was built in order to record

tomographic scans of the sample underload. The frame of the tensile testingdevice, made out of a PMMA tube,gave negligible absorption on the 2Dimages along the 180° rotation of thesample during the scan. Small double-shouldered tensile specimens with across-section of 1.5 * 1.5 mm2 and agage length of 5 mm were used. Thesample surfaces were mechanicallypolished using SiC paper and diamondpaste down to 1 µm. The tensiletests were carried out at roomtemperature using a constant crossheaddisplacement rate of 150 µm.min-1.Several scans were performed on thesame sample at the initial state andafter several increments of plasticdeformation. During the scan, theposition of the crosshead wasmaintained constant. Before and afterthe mechanical tests, the sample wasobserved in a Scanning ElectronMicroscope (SEM) operated at 20 keV.

RESULTS

Reconstructed 3D volumes showclearly the initiation of damage and itsevolution within the material. Internalpores, resulting from the manufacturingprocess, were observed in the material atthe initial state in the vicinity of the SiCparticles. At the surface, those pores hadbeen filled by the polishing process and,therefore, only a few of them weredetected in the SEM. Thanks to thephase contrast technique, cracks with anopening down to 0.5 µm - i.e. wellbelow the voxel size of 6.5 µm3 - couldbe detected in the reinforcing particles.

From a qualitative point of view, thedamage mechanisms observed in thebulk did not differ from those observed

Fig. 1: Threedimensional

reconstructionof some

reinforcingparticles in the

interior of anAl/SiC

composite.The average

size of theparticles is

120 µm.

X-RAY IMAGING

at the surface. Schematically, theevolution of damage as a function ofplastic strain can be described as follows:1 cracking of SiC particles,2 reinforcement/matrix decohesions,3 propagation of processing-inducedpores.

Representative images of theevolution of damage in the sample duringthe tensile test are presented in Figure 2.

From a quantitative point of view,however, the present experiment showsthat the number of cracked SiC particlesas a function of the plastic strain issubstantially larger in the bulk than atthe surface. Finite element modelling ofthe deformation process of a two-phasematerial is now being carried out to tryto account for the observed differencesbetween the bulk and the surface.

REFERENCES

[1] P. Cloetens, M. Pateyron Salomé,J.Y. Buffière, G. Peix, J. Baruchel, F. Peyrinand M. Schlenker J.Appl.Phys. 81 (1997) (9)133.[2] J-Y Buffière, E. Maire, C. Verdu,P. Cloetens, M. Pateyron, G. Peix, J. Baruchel,Mater. Sc. Eng. A234-236 (1997) 633.

WORKSHOP

ESRF - APRIL 1998

21

Fig. 2: Reconstructed images of the interior of the sample at the initial state (a) andafter two steps of plastic deformation (b and c). A tensile stress was applied alongthe vertical direction. Some cracks in the particles are indicated by white arrows.

a

b c

SASFLOW’ 98

ESRF-ILL WORKSHOP

«SOFT MATTER UNDER

FLOW AS PROBED BY

SMALL ANGLE

SCATTERING»

28-29 MAY 1998ESRF, GRENOBLE

FRANCE

DEADLINE

FOR REGISTRATION:10 MAY 1998

For further information,please contact Myriam Chakroun

Tel: +33/4 76 88 20 34Fax: +33/4 76 88 25 42

e-mail: chakroun@esrf.fr

X-RAY IMAGING

BRAGG DIFFRACTIONIMAGING OF MAGNETIC

CRYSTALS:

NEW RESULTS FROMNOVEL BEAMS

C. MEDRANO 1,2, I. MATSOULI 1,3,J. I. ESPESO 1,4, V. V. KVARDAKOV 5,

M. SCHLENKER 6, J. BARUCHEL 1

1 ESRF2 UNIV. ZARAGOZA, SPAIN

3 UNIV. WARWICK, UK4 UNIV. CANTABRIA, SANTANDER, SPAIN

5 KURCHATOV INSTITUTE, MOSCOW, RUSSIA

6 LAB. LOUIS NÉEL, CNRS, GRENOBLE, FRANCE

INTRODUCTION

Bragg diffraction x-ray imaging(x-ray topography) produces direct-space images of crystal defects,domains, or just thickness variations insingle crystals, via the inhomogeneousintensity of the Bragg-diffracted beams.

The advent of synchrotron radiationbrought high intensity making real-timeimaging possible and at high energymachines a large flux of high energyphotons, allowing the observation ofthick samples in transmission geometry.On third-generation machines, the smallemittance of the source again providesnew possibilities. Because imageblurring remains small even when thedetector is placed far from the sample,particularly at the long (145 m)beamline of ID19, the specimen-detector distance becomes a newadjustable parameter. It extends therange of options traditionally available:white beam vs monochromatic beamimaging, projection vs sectiontopography, selection of wavelengthand choice of the Bragg reflection.

We briefly present results obtainedat the ESRF, using beamline ID19,on magnetic materials: the oldestmagnetic material known, magnetiteFe3O4; hæmatite α-Fe2O3, manganesephosphide MnP, with a wealth ofmagnetic phases; and iron borateFeBO3. Although the scattering processused is Thomson (charge) scatteringand provides information on themagnetic moments only via the latticedistortion, quite spectacular resultswere obtained.

PROBING SYMMETRY: THELOW-TEMPERATURE PHASE OF

MAGNETITE

Magnetite, cubic above its Curiepoint and a ferrimagnet at roomtemperature, exhibits at 120 K atransition affecting its electric andmagnetic properties. Below this Verweytransition, the crystal symmetry is eithermonoclinic or even triclinic, implyingthat a single crystal splits into manydomains or twins. Magnetite is alsomagnetoelectric at low temperature,

with ferroelectricity coupled withferrimagnetism.

The domain structure was followedacross the transition, using white-beamtopography. The least complicatedsituation prevailed, as expected, whenthe transition was passed in a magneticfield favouring one of the easy <001>magnetization directions, or the nearesteasy directions [001]. Even then,however, so many domains weresuperimposed on any projectiontopograph that, contrary to standardusage, this provided no overall view.White beam section topographs, inwhich the effectively investigatedregion was a ribbon, some 20 µm thickacross the 0.8 mm thick sample,provided clues to the coexistingdomains, their shape and arrangement.

Figure 1a shows a typical result,corresponding to one Laue spot, to becompared with the single straight line asingle domain would yield. Unravelingthis data involved first thedetermination of matching lines,associated with adjacent domains in thespecimen; then the measurement of thegeometric splitting of thecorresponding lines associated with therelative rotation of the lattice planes fora number of Laue spots; andcomparison with calculation. Theoutcome is a plausible model, shown inFigure 1b, of monoclinic domains, withall walls satisfying the conditions ofneither magnetic nor electric chargedistributions, and no long-range elasticstress. Supplementary contrast isobserved, and is the first unambiguousevidence of extra distortions associatedwith triclinic symmetry.

The ability to vary the sample-to-film distance, to obtain high sensitivityto lattice plane rotation at large (40 cm)distances, and to confirm theassignment of matching elements atsmaller distances, proved essential.

WORKSHOP

ESRF - APRIL 1998

22

Fig. 1: a) Section topograph in whitebeam. Low temperature phase ofmagnetite. T = 110 K, B = 300 mT;593 reflections, λ = 0.3 Å, µt ≈ 1. b) Model for the monoclinic twins inthe sample.

Fig. 2: White beam

section topographon a 1.2 mm thick

hematite sample.21 reflection,

λ = 0.22 Å,µt ≈ 0.5.

Demagnetizedstate.

a

b

X-RAY IMAGING

AN UNUSUAL MAGNETIZATIONPROCESS: HÆMATITE

Hæmatite at room temperature is aweak ferromagnet, in which themagnetic moments are almostantiparallel and lie in the low anisotropybasal plane of the trigonal structure.Under an in-plane magnetic field,samples parallel to this (111) plane failto show domain walls. The white beamsection topography investigation of avery good crystal provides the clue tothis riddle (Figure 2): the walls arepinned along (111), and magnetizationoccurs through rotation of the magneticmoments. This is quite different fromthe standard process prevailing inmaterials with more sizeable anisotropy,where magnetization would proceedthrough the displacement of walls,which would furthermore be highlyunlikely to take this maximum areaorientation.

MAGNETIC PHASECOEXISTENCE: MANGANESE

PHOSPHIDE MNP

Orthorhombic MnP features amagnetic phase diagram withferromagnetic, helimagnetic, and fanstructures all within reach with moderatetemperatures and fields. Synchrotronradiation topography can reveal thecoexistence of two or three of themagnetic phases because each involvesslightly different lattice distortions. Inthe standard representation, the magneticphase diagram as an H,T plot showscoexistence lines. The experimental plotin the Ho,T plane, where Ho is theapplied field, shows bands instead of

lines (Figure 3a). This is related to thedifferent spontaneous magnetization ineach phase. A range of values of Hocorresponds to the same value of H =Ho+ Hdem because the demagnetizingfield Hdem changes continuously as oneof the coexisting phases grows. In thesame way, the triple point is spread intoa (small) area in the Ho,T plane.Figure 3b shows a projection topographobtained at this triple «point» and aschematic assignment of the variousparts of the image to the different phases.One of the noteworthy features is that thefan phase nucleates within the heli-ferromagnetic interface.

VISUALIZINGMAGNETOELASTIC

RESONANCES IN THE MHZRANGE: IRON BORATE

Iron borate is trigonal, isomorphouswith haematite and features strongmagnetoelastic coupling. White beamprojection topographs were recordedwhile a very good quality (111) singlecrystal plate, 50 µm thick, was excitedinto an elastic resonance by a 1.3 MHza.c. magnetic field superimposed on asmall perpendicular d.c. field, both in-plane. The vibration of the magneticmoments induces standing acousticwaves. In this imaging mode, thepredominant effect is that of latticeplane rotation of maximum amplitudeφo. A membrane type vibration leads tofocusing of different components of thebeam. Figure 4 shows some of theresults obtained from one Laue spot asthe excitation amplitude, hence therotation amplitude φo, was varied. Verysimilar pictures (ESRF Highlights 1996-

97) were obtained as a function of thesample-film distance L. This is inaccordance with the formal theoreticaldescription where the relevant parameteris the product Lφo, and again points tothe importance of the sample-to-detectordistance as a new free parameterprovided by the small emittance at theESRF. Such an experiment gives accessto the main characteristics of thevibration: amplitude, shape, polarization,wavelength and sound velocity. Inaddition this effect could be used toobtain a high frequency pulsed,«monochromatic», and focused x-raybeam.

WORKSHOP

ESRF - APRIL 1998

23

Fig. 3: a) Magnetic phasediagram, indicating the range

in applied field of thecoexistence on the videomonitor. b) White beam

topograph at the triple pointin MnP. 020 reflection.

Fig. 4: White beam topographs of thereflection in FeBO3 119 cm after the

sample as a function of the excitationamplitude; the resonance pattern

reveals the time integrated images ofthe pulsed, focused x-ray beam.

a

b

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

25

HIGH PRESSURE WORKSHOP

igh-pressure studiesin science, moreparticularly in the physicaland geophysical sciences,have greatly benefitedfrom the development ofnew techniques atsynchrotron radiationsources in the last 15years. The availabilityof very high-brillianceradiation at the ESRFsince 1993 has largelycontributed to thisdevelopment, and to thepopularization of the useof the pressure parameterin many fields, especially in those of structuralstudies and crystallography in extreme conditionsof pressure and temperature. With the setting up ofthe High-Pressure Commission of the InternationalUnion of Crystallography (IUCr) at the Seattlecongress in 1996, the success of the ESRF high-pressure program, and the first high-pressureexperiments at Spring-8 and APS in 1997, the timewas right to organize the first workshop of the newcommission. It was held at the ESRF from 21 to23 November 1997. 27 scientists representingthe main research groups active in these fieldsacross the globe accepted to give very interestingpresentations on the latest results anddevelopments at synchrotron radiation facilitiesand in associated fields. The program was dividedinto ten sessions. The first two were concernedwith non-structural techniques. They provided anoverview of the latest developments and results inmagnetic dichroism studies, Mössbauerspectroscopy, inelastic scattering, x-ray absorption,EXAFS of single crystals, and diffraction combinedwith spectroscopy. This was followed by twosessions on extreme conditions covering laser andresistive heating in angle- (ADX) and energy-dispersive (EDX) diffraction, extreme pressures,and structural studies of low-Z materials in theseconditions. After talks on liquids and large-volumepresses, attention was given to solving and refiningstructures with a comparison between neutronand x-ray techniques, a report of the work donerecently at the ESRF on nitrogen, and a provocativecomparison of the respective powers of ADX and

EDX. After an exciting session on strength, elasticityand kinetics, all hot topics at the moment as theanimated accompanying discussions showed, thelast two sessions concentrated on 2-D dataand detectors, and 3rd generation sources.The former covered 2-D data analysis, the«Fastscan» detector (see articles on pages 27and 30) and pixel detectors. The latter included areview of the ESRF highlights, a presentation of theSpring-8 high pressure facilities, and an overviewof the main technical and scientific projects in highpressure at APS, this including a proposal for a new‘multi-techniques’ sector dedicated to high pressureresearch. To round things off, an exciting forwardlooking view of the new science and techniquespointing over the horizon left much of theaudience with even bolder expectations. Theprogram also contained several very interestingdiscussion sessions which contributed greatly tothe success of the workshop. Over 90 participantsattended the meeting, including eleven youngscientists who benefited from travel awardsprovided by the IUCr. The organizers and HPCommission members thank the IUCr for thisfinancial support. Finally, I thank all those whocontributed to making this a great event, and animportant step in the development and successof the field of high pressure research. I cannotname them all here, but special thanks are dueto the HP Commission, and more particularlyto its chairman, Prof. R. Nelmes, for hisinvaluable help with the program.

D. Häusermann

H

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

26

LASER HEATING UNDERPRESSURE: A BRILLIANT

JOURNEY TOWARDS THECENTRE OF THE EARTH

G. FIQUET1, D. ANDRAULT2,A. DEWAELE1, T. CHARPIN2,

F. VISOCEKAS2, D. HÄUSERMANN3,M. KUNZ4 AND T. LEBIHAN3

1 LABORATOIRE DE SCIENCES DE LA TERRE,ECOLE NORMALE SUPÉRIEURE DE LYON

2 LABORATOIRE DE MINÉRALOGIE-CRISTALLOGRAPHIE,INSTITUT DE PHYSIQUE DU GLOBE, PARIS VI

3 ESRF, EXPERIMENTS DIVISION

4 LABORATORY OF CRYSTALLOGRAPHY, ETH ZENTRUM, ZÜRICH

INTRODUCTION

The study of the deep Earth has beenmotivating generations of scientists whohave to take up the challenge given bythe extreme conditions existing at thecentre of the Earth: over 300 GPa(3 million bar) and 5000 Kelvin. As thestudy of the propagation of elasticwaves created by earthquakes onlyallow the determination of the densityprofile of our planet, there remainsthe all important problem of thedetermination of the chemicalcomposition and crystalline structuresexisting in the deep Earth, as thesegovern the Earth global exchanges(thermal regimes, convection drifts,plate tectonics...). Thus relationships haveto be established between chemicalcomposition, crystalline structure andspecific volume over the whole range ofpressures and temperatures existingwithin our planet, and x-ray diffractionis by far the best technique to obtainreliable structural and molar volumedata on the compounds and materials ofinterest. Consequently, as very highpressures can only be generated inextremely small samples, the highbrilliance of the ESRF, coupled withadvances in detector, optics and highpressure technologies, has stimulatedthe rapid development of techniquesfor collecting structural data ongeophysical samples in extremeconditions of P and T. We report heresome recent results from a projectconcerned with in-situ x-ray diffractionstudies of laser-heated samples underpressure [1], and the examples selectedare studies devoted to the two majorconstituents of the deep Earth: iron andMgSiO3 perovskite.

EXPERIMENTALDEVELOPMENTS

Before this project, structural data onmaterials in laser-heated diamond-anvilcells (DACs) were obtained using white-beam energy-dispersive diffraction, atechnique which suffers from an intrinsiclow resolution and poor crystallitestatistics (hence unreliable intensity data)due to the small window of diffractionspace sampled, but thanks to the highbrilliance of the ESRF, it is now possibleto combine monochromatic angle-dispersive diffraction and image-platedetectors to collect quality data up topressures and temperatures in excess of90 GPa and 3000 K. This is achieved byfocusing the high brilliance beamproduced by two phased 40 mm periodundulators [2] using single-electrodebimorph mirrors [3] on the HighPressure beamline (ID30). The resultingfocal spot of about 8 µm x 15 µm(FWHM) is compatible with the size ofthe laser-heated hot spot, and thewavelength of the monochromatic beam,selected by a water-cooled channel-cutSi(111) monochromator in the 0.4 to0.5 Å range, is well matched to theaperture of custom-built DACs.Combining these beam characteristicswith an experimental set-up especiallydesigned for the project, and consistingof TEM00 CO2 and multimode YAGlasers, optical set-ups for on-line P,Tmeasurements and alignment of the

sample and beam, large aperture DACsallowing in-situ P,T measurements andfull 4θ data collection, for the first time ithas been possible to collect angle-dispersive diffraction data on image-plates during the laser-heating ofsamples only a few micron thick.

A NEW-PHASE OF IRON

Iron being the dominant constituentof the Earth’s core, information on itsbehavior at high P and T is fundamentalin Earth sciences, but despite numerousstudies on this subject [4] there is stillmuch uncertainty about its structure inthe P,T conditions relevant to the core.The accurate determination of the phasediagram of this element is indeed anexperimental challenge because of theextreme conditions involved, and evenbelow 100 GPa, recent x-ray diffractionexperiments have led to conflictingresults on the structure of its β-phase.Thus that region of the phase diagram,which was previously regarded assimple, is in fact complicated and clearlyin need of new experimental data.Indeed, as mentioned by Anderson [4],the final choice between the ε and γphases for the core depends on theoutcome of future studies aiming atproving the existence of the β-phase andidentifying its crystallographic structure.

The iron phase diagram up to 100GPa and 2700 K has thus been studiedusing the best diffraction technique

6 8 10 12 14 16

2-θ[°]

Inte

nsit

ies

(a.u

.)

18 20 22 24

Iron + Corundum, P ~ 45 GPa, T ~ 2125 K

Fig. 1: Full structure refinement of a spectrum recorded at 2125 K, 44.6 GPa.The cell is orthorhombic with space group Pbcm and iron location {0.239, 0.472,0.25}. Observed, calculated and difference spectra are shown, with lower ticks foriron and upper ticks for Al2O3.

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

27

currently available, and these resultsintroduce strong constraints on itsstructures at high P and T. Havingsignificantly improved both the resolutionand the reliability of the data, it has beenpossible for the first time to perform fullstructural Rietveld-refinement in theseextreme conditions of P and T (Figure 1).The space group determined is Pbcm, andthe atomic topology is close to thatof the ε-hcp phase. The structure isalso closely related to the lower pressure,high-T polymorph, γ-fcc. The high-Tpolymorph appears unquenchable atmoderate pressures, but the spectra ofthe back-transformed γ-phase showsome anomalies which can explainambiguities reported in previous structuredeterminations [5].

To eliminate possible artifactsintroduced by the pressure transmittingmedium (corundum, which is also used asthermal insulator), the occurrence of theorthorhombic lattice in a SiO2 mediumwas also checked. This is illustrated inFigure 2 which shows results obtained upto about 100 GPa. Unfortunalely at thispressure the sample was not sufficientlyinsulated from the diamonds, and it wastoo difficult to obtain in-situ spectraduring stable laser heating. However,spectra quenched from about 2500 Kclearly show the doubling of the 100 and101 lines of the hcp-lattice, here again anevidence of a transformation at high P andT. All the experimental lines are explainedby an orthorhombic lattice similar to thatpreviously observed, and in contrast withresults obtained at moderate pressures, thestructure of the high-T polymorph is nowpreserved after the quench. At thispressure, the orthorhombic lattice is foundto be about 1% denser than ε-iron.

THE STRUCTURE OF MGSIO3PEROVSKITE

The perovskite form of (Mg,Fe)SiO3being currently accepted as the dominantphase of the Earth’s lower mantle (700 to2900 km deep), its equation of state(EOS) plays an important role in manyfields of geophysics. It is howeverpresently impossible to choose betweenthe perovskite-pure and perovskite-magnesiowüstite (Mg, Fe)O models forthe Earth’s lower mantle on the basis ofthe existing data, and in-situ high P and Tdiffraction is certainly the only methodavailable to measure correctly its EOSand solve the stuctural problem. Previousstudies were conducted in the stabilityfield of the perovskite, but energy-

dispersive diffraction and large-volumepresses limited the performance and P,Tranges to 30 GPa and 2000 Krespectively. Using the techniquedescribed earlier, our measurements onMgSiO3 perovskite were extended to86 GPa and 2700 K. Here however thenew on-line image-plate detector (the«Fastscan» [6]) now available on the ID30beamline was used for the data collection.

Silicate perovskite MgSiO3 sampleswere synthesized from syntheticMgSiO3 enstatite crystals or syntheticMgSiO3 glass mixed with platinumpowder, and once loaded in a largeaperture DAC, the starting materialswere transformed at high P using eitherthe CO2 or the YAG infrared lasers,

depending of the pressure transmittingmedium. The temperature wasdetermined by analyzing the thermalemission spectra recorded during thediffraction measurements and thepressure conditions were calculatedfrom the EOS of platinum, used here asinternal pressure calibrant [7].

Le Bail profile refinements wereperformed on the diffraction patterns toobtain reliable high P, high T cellparameters for the sample and thepressure calibrant up to 86 GPa and 2700K, and the most remarkable result wasthat Rietveld structural refinements weresuccessfully carried out on selectedpatterns in these extreme conditions [8](Figure 3). This gave for the first time

1.2 1.4 1.6 1.8 2 2.2

Distance (Å)

Inte

nsit

y (a

.u.)

110102

101002

100

Sti

Sti

StiSti

200/122112/022

111

021

002

110

020

Sti

StiStiSti

hcp-iron

Ortho-iron

P = 100 GPaFig. 2:

Diffraction spectraof hcp and

orthorhombic ironrecorded at 100 GPa

in a SiO2 medium(labeled Sti).

The 101 reflection ofε-iron is truncated

for better clarity. Thetop spectrum,

quenched from about2500 K, clearly

shows the doublingof the 100 and 101

ε-iron lines, evidenceof the phase

transition toward theorthorhombic phase.

6 8 10 12 14 16

2-θ[°]

Cou

nts

18 20

Fig. 3: Full Rietveld structure refinement of a diffraction spectrum of MgSiO3 at86 GPa and 2310 K integrated from an image plate exposed for 10 minutes using

a monochromatic beam focused to 10 µm x 20 µm. Sample and platimum (thepressure calibrant) reflections are shown by lower and upper ticks respectively.

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

28

precious structural information on thesecompounds, as for instance the firstobservation of the increase of the internaldistortion of the SiO6 octahedra withincreasing pressure in a powder sample.Furthermore, the data analysis allowed usto identify a set of thermoelasticparameters to constrain the compositionalmodel of the Earth’s lower mantle.Assuming that the thermoelasticparameters obtained from this study areapplicable to perovskites with moderateiron content, then the comparison of thedensity and KT profiles calculatedfor a mixture of perovskite andmagnesiowüstite with those obtainedfrom the PREM [9] model indicates thata pure perovskite lower mantle is veryunlikely. On the other hand, a very goodmatch between the PREM density

and KT profiles is obtained for amixture of 83 vol% (Mg0.93, Fe0.07)SiO3perovskite and 17 vol% (Mg0.79Fe0.21)Omagnesiowüstite [8].

REFERENCES

[1] G. Fiquet, D. Andrault, J.P. Itié, Ph. Gillet,and P. Richet (1996) High-pressure and high-temperature x-ray diffraction study of periclasein a laser-heated diamond anvil cell. Phys.Earth Planet. Int., 95, 1-17.[2] J. Chavanne, P. Elleaume and P. VanVaerenbergh, Phasing multi-segment undulators,ESRF Newsletter No 25, 1996, 12-14.[3] J. Susini, Active/Adaptive optics at theESRF, ESRF Newsletter No 21,1994, 20-23.[4] O.L. Anderson (1995) Mineral physics ofiron and the core. Reviews of Geophysics,Supplement: 429-441. U.S. National report tointernational union of geodesy and geophysics1991-1994.[5] D. Andrault, G. Fiquet, M. Kunz,F. Visocekas and D. Häusermann (1997) Theorthorhombic structure of iron: an in-situ studyat high-T and high-P, Science, Vol. 278, pp. 831.[6] A new image plate detector system fordiffraction experiments at high pressure, ESRFHighlights 1997-1998, 93-94.[7] D. Andrault, G. Fiquet, J.P Itié, P. Richet,P Gillet, D. Haüsermann and M. Hanfland.(1997) Pressure in the laser-heated diamondanvil cell: implication for the study of theEarth’s interior. Eur. J. Mineral., in press.[8] G Fiquet, D. Andrault, A. Dewaele,T. Charpin, M. Kunz and D. Haüsermann.(1997) P-V-T equation of state of MgSiO3perovskite. Phys. Earth Planet. Int., in press.[9] A.M. Dziewonski and D.L. Anderson(1981), Preliminary reference Earth model,Phys. Earth Planet. Int., 25: 297-356.

DENSITYMEASUREMENTS OF

LIQUID IRON ALLOYS ATHIGH PRESSURES:

TOWARDS A BETTERUNDERSTANDING OF

THE PLANETSC. SANLOUP1, I. MARTINEZ2, F. GUYOT3,P. GILLET1, G. FIQUET1, M. MEZOUAR4,

D. HÄUSERMANN4 AND T. LEBIHAN4

1 LABORATOIRE DE SCIENCES DE LA TERRE,ECOLE NORMALE SUPÉRIEURE DE LYON

2 LABORATOIRE ISOTOPES STABLES, INSTITUT DE

PHYSIQUE DU GLOBE, PARIS

3 LABORATOIRE DE MINÉRALOGIE-CRISTALLOGRAPHIE,INSTITUT DE PHYSIQUE DU GLOBE, PARIS

4 ESRF HIGH PRESSURE GROUP, GRENOBLE

Physical properties of iron-basedliquids are of much interest to betterunderstand both the current state ofplanetary cores and their formationduring the differentiation of planets.Here we present the first experimentsperformed on metallic liquids in the Fe-Ni-S system, which might be relevant atleast to the terrestrial outer-core and themartian core. Using a large-volumepress apparatus (a Paris-Edinburghpress), the P-T range of 0-4 GPa and20-1250 °C was explored by measuringthe absorption profiles, hence density,of samples using high-energy x-rays.Equations of state of liquid ironalloys are therefore on the way to bedetermined, along with accuratemelting-phase diagrams as a function ofpressure and temperature relevant togeophysical conditions.

GEOPHYSICAL INTERESTS

Density measurements of Fe-basedliquids at pressure and temperaturerelevant to planetary cores are essential tomodel accurately the core compositionand convection. This should help resolvetwo important geophysical issues: thegeneration of the Earth's magnetic fieldand the thermal history of the planet.

Also relevant to these measurementsis the differentiation of planets, i.e. at firstorder, the individualization of a metalliccore towards the center of the planet. Allthese phenomena refer to the liquid stateof core materials, which concerns atleast the outer terrestrial core, but alsoprobably Mars, Venus and some Galileansatellites such as Ganymede for example,as a substantial magnetic field (roughly a

ACKNOWLEDGEMENTS

We acknowledge the use ofexperimental facilities provided bythe Extreme Conditions Consortium(«ECC», a collaboration between theUniversities of Paris VII, Lausanne,Uppsala, the Ecole NormaleSupérieure de Lyon and the ESRFHigh Pressure Group). We aregrateful for the support and efforts ofS. Bauchau and M. Hanfland from theESRF, J.P. Itié from Paris VI / LURE,F. Guyot and P. Richet from Paris VII,and P. Gillet from the ENS Lyon. The«Fastscan» detector project is acollaboration between the ESRF andthe University of Erlangen, Germany.We thank M. Thoms for all his workand A. Winnaker for supporting thisproject.

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

29

tenth of the Earth's field in intensity)was recently detected by the NASA'sGalileo spacecraft.

Seismic waves, generated byearthquakes, have revealed to be powerfulprobes of density and elastic properties ofthe interiors of the Earth, and soon of theMoon and Mars. Indeed, seismic datapredict that the terrestrial core iscomposed of an iron-nickel alloy up to90 %, plus 10 % of a light element, whichis mainly supposed to be either oxygen,silicon, sulfur, or carbon, or a mixtureof these elements. The composition ofthe core can also be assessed usinggeochemical models of planetaryformation, and in fact the consensuswhich emerges now is that sulfur seems tobe quite an ubiquitous component ofmetallic cores, but in varying proportions(up to 25 wt % for the Martian core forexample, and less than 4 wt % in theEarth's core).

The P-T domain of interest differsfrom one planet to another, ranging fromas high as 130-330 GPa and 4500-5500 Kfor the Earth, to 8-12 GPa and 1500-2000 K for Ganymede. In theexperiments discussed below, we havereached a P-T range of 5 GPa and1200 °C, which is already of great interestfor such geophysical and astrophysicalproblems, even though it needs to beextended to more realistic conditions.

PRELIMINARY RESULTS

Using a large-volume pressapparatus (known as the Paris-Edinburgh press - see Figure 1), the P-Trange of 0-4 GPa and 20-1250 °C was

explored, and high-energy x-rayabsorption data (λ = 0.21 Å) werecollected in-situ using the methoddeveloped by Katayama et al. [1] for thestudy of liquid tellurium. Thesemeasurements were carried out on theESRF High Pressure beamline (ID30)using radiation from the 70 mm periodwiggler. The pressure and temperatureconditions were determined bycomputing the intersections of theisochoric lines for hexagonal BN (thesample is contained in a hBN cylinder)and γFe, just before melting. Thetemperature was also calibrated againstthe power delivered by the system usingthese isochoric lines, which allowed usto determine an empirical law for T as afunction of the input power, and then toextrapolate it to higher temperatureswhen the Fe-S alloy was molten.

Disappearance of crystallinediffraction peaks allowed a precisedetermination of the melting point ofsamples, while their density wasobtained from the x-ray absorptioncurve (Figure 2) as x-ray absorptionobeys the Beer-Lambert law:

I/Io = ∫beam size exp (–µliqρliq(X)–µenvρenv) dx

where I is the intensity of the x-raybeam, µ the mass-absorption coefficient,ρ the density, and 'env' stands for thesample environment.

Fig. 1: The Paris-Edinburgh press (modelV4). Using a force of300 tonnes, 300 mm3 ofmaterial (gasket, heater,calibrant, sample...) arecompressed so that themaximum pressure ofthe central 100 mm3

of sample is 12 GPa.The device weighsapproximately 50 kgand measures 20 cm by20 cm by 25 cm. Withhigh-resistivity graphiteheaters, the sampletemperature is raisedto 1600 °C by passinga current of 100 Athrough the anvils,giving a maximumpower of 250 W.

Fig. 2: X-ray absorptionscans of 73 % Fe-27 % Sliquids. The logarithm ofI/Io is fitted according to

the Beer-Lambert law(continuous lines, see

text) at different pressures(1.5, 3.0 and 3.9 GPa),

and around 1250 °C; thebackground signal

corresponds to the sampleenvironment.

HIGH PRESSURE WORKSHOP

ESRF - APRIL 1998

30

During our first set ofexperiments, liquids were obtained inthe Fe-S system by mixing Fe and FeSpowders to have 25wt % of sulphur.It appears that the presence ofsulfur dramatically reduces the bulkmodulus, Ko, as Nasch et al. [2]already noticed when carrying out anultrasonic interferometry investigationof molten Fe-5 %Ni-10 %S at ambientpressure. Indeed, they obtained avalue of 63 GPa for Ko, while it is110 GPa for pure liquid iron, andthe value we obtained in theseexperiments is as low as 27 GPa with2.5 times their quantity of sulfur.This result could be of considerablegeophysical and astrophysicalimportance, if confirmed in ourfuture studies. In particular, it couldbe used to determine the maximumsulfur content in the Earth's core, andsoon in the Martian core, by directcomparison with core bulk moduliderived from seismological data.

As mentioned earlier, absorptionand diffraction data were collectedsimultaneously, so that in addition tothe observation of the disappearanceof the crystalline state, the diffractiondata gave a first order approximationof the radial distribution function.The integrated diffraction data (the«Fastscan» detector was also usedhere to collect 2-D images - see articleby Fiquet et al. on page 26) of liquid73%Fe-27%S at 3.9 GPa and 1235 °C(Figure 3) show a broad peak ataround 5.5° in 2θ, which correspondsto an interatomic distance of 2.1 Å forthe position of the first neighbours,and we can also distinguish a second

weaker peak at about 10.5° in 2θ. Lastly, and of great importance, is

the high-precision measurement of theequations of states of the products ofmetal oxidization. This is carried outusing the same high-pressure device,but in a conventional diffractionmode, and also with image plates.Here again the «Fastscan» detector isused, and an example of integratedspectra is shown in Figure 4. Imagesare collected by first illuminating the

sample and its surroundings, then onlyits surroundings by translation of thewhole press, and the differencespectra obtained are of high enoughquality to allow analysis by Rietveldrefinement. This opens up thepossibility of studying the equilibriabetween liquid metals and solidoxides along planetary P-T profiles.These equilibria are the first orderparameters of the establishmentand evolution of redox states inthe planetary interiors, and theseeventually determine the nature of thefluids at the surfaces. Considering thepromise of these preliminary results,we are now planning to explore theFe-Ni-S-O system at higher pressures(up to 8-10GPa) and for various othercompositions.

REFERENCES

[1] Y. Katayama (1996) High Pressure Research,14, 383-391.[2] P.M. Nash (1997) Science, 277, 219-221.

0

2-θ angle [°]

Inte

nsit

y (a

.u.)

5 10 15

25

15

5

– 5

Fig. 3: 1-Dintegrateddiffraction spectraof 73 % Fe-27 % Sliquid at 3.9 GPaand 1235 °C. Theyare fitted with apolynomialfunction whichshows the first andsecond rings of theradial distributionfunction.

2

2-θ [°]

Inte

nsit

y (a

.u.)

4 6 8 10 12 14

ACKNOWLEDGEMENTS

We are grateful to S. Bauchau for histechnical work, and to all the ESRFsupport staff which contributed to thesuccess of these first experiments.

Fig. 4:Diffraction

spectra of Fe2O3integrated from2-D «Fastscan»images obtained

in 20 secondsusing unfocussed

high-energyradiation. Top:

sample, calibrantand gasket,

middle: calibrantand gasket,

bottom:recovered Fe2O3

spectrumsuitable for

Rietveld analysis.

ESRF - APRIL 1998

31

A EUROCONFERENCE AND NEA WORKSHOP

ON SPECIATION, TECHNIQUES, AND FACILITIES FOR RADIOACTIVE

MATERIALS AT SYNCHROTRON LIGHT SOURCES

will be held at the ESRF, Grenoble,4-6 October 1998

This meeting will introducethe applications of synchrotron-based analytical techniques toenvironmental and radionuclidescientists, provide a forum forteaching and scientific discussion,and establish a possible cooperative

scientific network. Invited speakerswill give introductory lectures onthe types of information that canbe obtained from synchrotron-basedtechniques and report the latestresults on radionuclide work.

Persons who are interested in

applying synchrotron radiation canobtain necessary information fromlectures on protocols which are inplace for radionuclide research atseveral synchrotron storage ringsthroughout the world and duringa tour of the ESRF.

For further information,please refer to the conference home page at

http://www.fz-rossendorf.de/An-XAS-98

or send e-mail to T.Reich@fz-rossendorf.de.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

32

n 1988 the ESRF had its first discussions withexternal users to consider how an increase in thenumber of beamlines at the facility might beachieved. The ESRF had been designed andfunded as a facility which would have essentiallyinsertion device ( ID) x-ray sources, but it wasclear that scientists from the member countrieswho had no access or insufficient access tobeamlines at National Synchrotron RadiationFacilities could benefit from the use of theradiation generated by the bending magnets(BM) in the storage ring. These sources, althoughless intense than the ID sources can, withappropriate focusing optics, produce x-rayintensities at the sample position which exceedthat found in many national facilities. TheCouncil agreed at its June 1990 meeting thatscientists from the member countries could formgroups to exploit these BM sources provided thatthis did not divert either manpower or fundsfrom the main ESRF program. These groups arecalled Collaborating Research Groups (CRG).Several groups were formed in a very shortperiod of time, and their scientific programstogether with beamline designs were submittedto the Science Advisory Committee (SAC) forformal approval in 1991 (see Newsletter Nr 9).Formal contracts with the ESRF were signed soonafterwards and construction of the first CRGbeamlines in the experimental hall started in1992. The ESRF supports the CRG program, andprovides x-rays free of charge in return for use ofthe instrumentation on the CRG beamlines for1/3 of the scheduled beam time. The ESRF alsoprovides the front-ends which transport the x-rays from the source to the beamlines.

In the future the cost of the ESRF front-endswill be refunded by the CRG.The CRG program is now well established, with4 CRGs having operated a user program sincethe end of 1994, and with 4 more groups havingbeamlines in an advanced stage ofcommissioning. A 9th CRG was approved by theCouncil at its June 1997 meeting, and the ESRFis aware that three possible further projects areunder discussion. In view of the burden on thegeneral ESRF infrastructure by further beamlines,the ESRF Council decided at its December 1997meeting to set 16 independent CRG beamlinebranches as the threshold for the next review ofthe situation by the Council. The scientific aimsof each of the first 9 CRGs are given in Table 1.The GRAAL experiment at the ESRF also operatesas a CRG although it is dedicated to nuclearphysics measurements, and does not usesynchrotron radiation (see more detailson page 41).The user operation of the ESRF started inSeptember 1994, and the shifts (8-hour runningperiods) allocated for user experiments on theoperational CRG beamlines, up to the end of1997, are shown in Figure 1. In total, 1723 shiftshave been carried out for public users and 4014shifts have been allocated to the CRGs «private»users. The total of 5737 compares with 18029shifts being available for public users on theESRF beamlines during the same period. Usingthese scheduling figures as a measure, we seethat the public user program has been increasedby nearly 10% due to the presence of the CRGs,and that the total scientific activity at the ESRFwas increased by over 30% during the four-year

COLLABORATING RESEARCH GROUPS: A REVIEW

I

CRG NATIONALITY STUDIES OPERATION DATESSINCE FORESEEN

BM1 (SNBL) Swiss/Norwegian Multipurpose Jan. 95

BM2 (D2AM) French Materials / Biology Sept. 94

BM7 (GRAAL) Italian Photoproduction of particles Mid. 95

BM8 (GILDA) Italian Multipurpose Sept. 94

BM20 (ROBL) German Materials / Radiochemistry Sept. 98

BM25 (SPLINE) Spanish Multipurpose 2001

BM26 (DUBBLE) Dutch / Belgian Multipurpose Sept. 98

BM28 (XMAS) British Magnetic scattering April 98

BM30 (FIP) French Proteins structure Sept. 98

BM32 (IF) French Interface studies (multipurpose) Sept. 94

Table 1

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

33

start-up period. It is expected that the CRGs willbe able to maintain a significant contribution tothe scientific life of the ESRF as more beamlinesbecome operational during 1998.This extra activity is of course not obtained «freeof charge», the total extra-capital investment bythe countries having one or more CRG beamlineswill be about 200 MFF by the time the ten groupsall have operational beamlines. In addition, thesalaries of beamline staff and the travel costs forabout 1350 CRG users who have come to theESRF so far, to carry out 450 CRG proposals, arefunded from the CRGs' own budgets.

CRG infrastructure - ESRF/CRG collaborationsAt the inception of the CRG program, the CRGsformed a «club» as a forum for sharing newideas, for discussing the difficulties involved inworking in a «foreign» laboratory, and as ameans of communicating with the ESRF. Thisquickly led to the ESRF Council’s agreement tothe appointment of a CRG liaison engineer to beresponsible for advising the CRGs on all technicalmatters relating to the installation and operationof their beamlines. When the first beamlinesbecame operational, the CRGs started to needhelp with the internal ESRF administration of theirown user programs and collectively providedfunding for employment of secretarial assistance.This user oriented work has now increased andcurrently two part-time secretaries are employed.During the last year, CRGs have also beencollectively funding a technician post to assistwith various construction and user-oriented taskson their beamlines. These staff are employedby the ESRF on behalf of the CRGs.

The ESRF and the CRGs have co-operated in two more significantareas. The first follows aninitiative by the Swiss/NorwegianCRG to split the 6 mrad bendingmagnet radiation fan into two toenable two beamlines to operateindependently. Other groupsasked for a 9 mrad wideradiation fan to be delivered tothe experimental hall. The ESRFsubsequently redesigned part ofthe front-end so that this couldbe achieved, and this

development has greatly eased the spaceproblems associated with the installation of twosets of optical components in the confined spaceavailable in a single optics hutch. Three CRGshave now chosen this 9 mrad option, despitehaving to pay for the additional front-end costsinvolved.The second area of co-operation has been theprovision of office and laboratoryaccommodation for the CRG staff and users. Aspreviously stated, the original ESRF plans did notinclude any provisions for the CRGs. Throughthe CRG Club, the CRGs jointly expressed aneed for accommodation «as close as possible tothe beamlines». This is why there are now somebuildings projecting outside the experimentalhall. The ESRF Technical Services Division madethe planning and construction of these buildingspossible, while the ESRF Administration togetherwith the CRGs made long-term financialagreements ensuring there would be no costto the ESRF.With the move from the construction phase ofthe ESRF towards routine operation, the CRGsare, through the CRG Club, extending the co-operation between themselves, and wherepossible, with the ESRF. Since the start of theCRG program, the ESRF has recommended thatthe CRGs adopt ESRF technical solutions wherepossible as such a policy has many advantages.The CRGs are now building up a stock of spareparts which are commonly available, and it ishoped that in the near future this concept will beextended to other areas, e.g. participation in adetector pool at the ESRF.

I. Kilvington

2000

1800

1600

1400

1200

1000

800

600

400

200

01994 1995 1996 1997

CRG shiftsESRF shifts

Fig. 1: shifts onCRG beamlines.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

34

A NEW TECHNIQUE:PHYSICALLY ESTIMATED

TRIPLET PHASES R.H. MATHIESEN1, F. MO1, A. EIKENES1,

T. NYBORG1 AND H.B. LARSEN2

1 DEPT. OF PHYSICS, NTNU, TRONDHEIM

(NORWAY)2 STAVANGER COLLEGE, STAVANGER (NORWAY)

Information on phase is lost in thegeneral diffraction experiment with hardx-rays, a fact which gives rise to the so-called phase problem in crystallography.This is in contrast to diffraction ofvisible light or electrons, where thescattered beams can be collected bysuitable lenses, to provide directly areconstruction of the scattering object.However, even with hard x-rays, phaseinformation can be retrieved fromspecially designed experiments.

Electron diffractionists have knownfor about 50 years that phaseinformation is available as interferenceeffects between several beamsdiffracted simultaneously in perfectcrystals (n-beam diffraction) [1, 2].

Experimental evidence was obtainedlater with x-rays [3]. This direct,physical approach to structure-factorphase acquisition became of muchgreater practical importance whenseveral groups [4, 5, 6] in the early1980’s found that phase effects can besignificant also with truly mosaiccrystals. Mosaic, or non-perfect crystalsare vastly more abundant than theperfect specimens.

STATUSTheoretical descriptions of n-beam

x-ray diffraction have been developedby several authors on different levels ofapproximation; for a review seeWeckert and Hümmer [7]. In a recentcontribution Thorkildsen and Larsenhave considered the influence ofvarious parameters describing theradiation and its interaction with afinite, non-perfect diffracting crystal, onthe solution of the Takagi equationsfor three-beam diffraction [8]. Theexperimental side has grown verystrongly over the past decade, first andforemost through the diligent work of

Weckert, Hümmer and collaboratorsin developing highly sophisticatedinstrumentation and measurementtechniques. Triplet phases can now beestimated physically even for crystals ofmacromolecules. At present, phaseshave been acquired successfully for atleast 5-6 different proteins.

EXPERIMENTThe n-beam experiment in its simplest

form comprises three beams (n = 3), oneincoming and two diffracted beams.A very simple representation is shownin Figure 1. In the crystal the incidentbeam K0 can be scattered into the KHdirection by two different pathways,either by the direct wave diffracted at thelattice planes H, or by the Umweg wavediffracted first at lattice planes L, and thenat the H-L planes. In analogy withholography experiments the totalwavefield in direction KH will dependboth on the amplitudes of the componentwaves - FH, FL and FH-L - and the phasedifference D = (fL + fH-L) - fH. The latterquantity can be rewritten as a phase sumF3 = f-H + fL + fH-L, a so-called three-

SNBL (SWISS-NORWEGIAN BEAMLINE) ON BM1

The Swiss-Norwegian facility is split into two branch lines:

one dedicated to single-crystal diffraction (2.5 mrad fan of radiation, focused beam, high-resolution

monochromator) and the other to powder diffraction, EXAFS and topography (1 mrad fan of

radiation, unfocused, channel-cut monochromator). To illustrate the kind of research carried out on the

two beamlines, we present and discuss one recent investigation from each branch line.

From left to right:P. Pattison,K. Knudsen,H. Emerich,W. Van Beek;in the car: H.P.Weber.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

35

phase structure invariant, which is aphysically unique quantity of the crystal.In the experiment the crystal is rotatedabout the normal to lattice planes H, suchthat the planes L are broughtsuccessively into and out of thediffraction position, while reflection Hremains fully excited. During thisrotation, called a Y-scan, the primarydiffracted intensity IH is recorded. Theinterference between the diffractedbeams which contains information aboutthe triplet phase F3, is projected out asa characteristic modulation of theintensity profile. In addition to thisphase dependent «signature», theprimary diffracted intensity may beenhanced (Umweganregung) or depleted(Aufhellung) by phase independentcontributions from the other beams.

APPLICATIONSPhysically estimated triplet phases

(PETP) have several interestingapplications, of which only a fewhave been realized so far. Anobvious application is to initiate thedetermination of molecular structures.Apparently, only one unknown smallstructure has been solved from PETP[9]. Feasibility studies have shown thatsolution of macromolecular structures iswithin reach. Other applications includedetermination of absolute structure andstudies of quasicrystals [7].

A NEW APPLICATIONPhases are much more sensitive to

small changes in structure than are thestructure-factor amplitudes that can bederived from the intensities measured inthe standard diffraction experiment.Therefore, PETP could presumably beused as well to discriminate betweenseveral possible and closely-relatedmodels of a structure. This novelapplication was explored in a study of

the complex a-D-glucose ∑NaCl ∑H2O(6:3:3) in space group P31, with unit-cellvolume V = 4180 Å3. A structure ofthe complex has been published(Model CA) [10]. However, there existsone alternative solution (Model GE)[11]. We found from refinements thatthe primary difference in structure isan interchange of the Cl and waterO positions, which is accompaniedby smaller, mainly translatorydisplacements of glucose rings, and areorientation of water molecules and oneof the glucose OH groups. This leads toa reversal of the polarity in chains of H-bonds along the polar c axis. Thechanges in structure imply significantchanges in phase for about 50% of thestructure factors. In contrast, tripletphases being sensitive to the model areextremely scarce. Thus, only about0.05% of the calculated three-phaseinvariants with amplitudes suitable forphase measurement had a triplet-phasedifference |DF3| > 30o! By measuringseveral model-sensitive three-phaseinvariants and comparing with thosecalculated for the refined models CA andGE we expected initially to be able to

identify one correct model of the pair.

EXPERIMENTALThirteen crystals, all cut from one

large single crystal of the complex weresubjected to physical phase estimationwhich involved studies of model-sensitive triplet phases. Each phaseassignment was based on the intensityprofiles mapped out in Y-scans of thepair of triplets -H/L/H-L and H/-L/-H+L,corresponding to phases +F3 and -F3,respectively. Two distinct right-handedcell matrices are possible in this spacegroup. It was ascertained that the choiceof unit cell of the crystals was internallyconsistent and in agreement with theindexing of the native data sets CA andGE. In total 309 pairs of three-beaminterference profiles were collected forthe thirteen crystals. The measurementsincluded 89 different triplets.

RESULTSThe estimated triplet phases have

been compared with the parent valuescalculated for the refined structuremodels CA and GE. A statisticalanalysis provides highly significantevidence that two structures are present,but do not coexist in the same crystal.Six crystals are in agreement withmodel CA, seven crystals fit the GEmodel. The estimated random error inthe phase assignment is 19.7o; for allcrystals the model with the largest meanvariance can be rejected at asignificance level p << 0.001. Theresults are presented in Figure 2. Thecell parameters of the two structures areequal within 1 esd., as determined withsynchrotron radiation, λ = 1.0000 Å.

H

Ψ

LH-L

KO

KL

KHFig. 1: A three-beamdiffraction caseillustrated in realspace. K0 is theincoming beam, KHand KL are theprimary and thesecondary diffractedbeams, respectively.

0

Model CA = X (deg.)

Mod

el G

E =

Y (

deg.

)

X = < (Φ – Φ )2 > 1/2est3

calc3,CA

Y = < (Φ – Φ )2 > 1/2est3

calc3,GE

CA

GE

20 40 60 80 100 120 1 6

100

80

60

40

20

0

Fig. 2:Classification

regions given forthe mean errors

calculated relativeto the two models.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

36

As an example, the intensity profilesfor two triplets, one model insensitivewith |DF3|~ 20o (a), and one modelsensitive with |DF3| ~ 170o (b) , areshown in Figure 3 for both structuremodels. In each pair, the upper profilecorresponds to the triplet -H/L/H-L. Themodified intensity is plotted relative tothe intensity at the two-beam level, IH =1.0, as a function of Y. Note the invertedasymmetry and the reversal of high/lowamplitude from top to bottom profilebetween the two triplet pairs of Figure3(b), in accord with phase assignmentsdiffering by about 160o. A full accountof this work will be published [12].

The Y-scans were carried out on a 6-circle diffractometer (Prof. Hümmer,Univ. of Karlsruhe) located on theSwiss-Norwegian Beamline.

REFERENCES

[1] E. Fues, Z. Phys. 125 (1949), 531 - 538.[2] (a). S. Miyake and K. Kambe, Aca Crystallogr.7 (1954) 220. (b). K. Kambe, Acta Crystallogr. 7 (1954) 777 - 779. (c). K. Kambe,J. Phys. Soc. Jpn. 12 (1957) 13 - 31.[3] M. Hart and A.R. Lang, Phys. Rev. Lett. 7(1961) 120 - 121.[4] S.-L. Chang, Phys. Rev. Lett. 48 (1982) 163 -166. [5] (a). B. Post, Acta Crystallogr. A39 (1983) 711- 718. (b). P.P. Gong and B. Post, Acta Crystallogr.A39 (1983) 719 - 724.[6] (a) G. Thorkildsen and F. Mo, Abstracts ECM-7, Jerusalem (1982) p. 6. (b) G. Thorkildsen andF. Mo, Abstracts ECM-8, Liège (1983) p. 258.[7] E. Weckert and K. Hümmer, Acta Crystallogr.A53 (1997) 108 - 143.[8] G. Thorkildsen and H.B. Larsen, ActaCrystallogr. A54 (1998), 120 - 128; ibid. 129 - 136

ibid. 137 - 145.[9] F. Mo, B.C. Hauback and G. Thorkildsen,Acta Chem. Scand. A42 (1988) 130 - 138.[10] G. Ferguson, B. Kaitner, B.E. Connett andF. Rendle, Acta Crystallogr. B47 (1991) 479 - 484.[11] R. Fröhlich, Personal communication(1989).[12] R.H. Mathiesen, F. Mo, A. Eikenes,T. Nyborg and H.B. Larsen, Acta Crystallogr.A54 , in press.

EXAFS: HYDRIDETREATMENT CATALYSTS

T. SHIDO1, R. PRINS1AND H.P. WEBER2

1 TECHNISCHE CHEMIE, ETHZ, ZURICH

(SWITZERLAND)2 PHYSICS DEPT, UNIV. OF LAUSANNE, DORIGNY

(SWITZERLAND)

EXAFS is the only method suitableto determine the local structures of bulk,non-crystalline materials under arbitraryatmosphere temperature and pressureconditions. From an experimental pointof view, EXAFS is a particularly simpletechnique; essentially, one measures x-ray absorption spectra of the samples ofinterest as a function of energy; the skillis mostly in the analysis of the spectra,and progress has come mostly fromcomputer-intensive modelling. Thetechnique has been widely used toinvestigate the structure of supportedcatalysts. Up to recently, only thecontribution from the first shell of atoms(mostly M-O) to the EXAFS spectrawas analyzed; this yielded thecoordination number of the absorberatom. However, with the recentdevelopment of multiple scatteringtheory, the analysis of the contributionfrom the higher shells (mostly M-Minteraction) has become possible; it

a

b

ACKNOWLEDGEMENTS

We are greatly indebted toE. Weckert, K. Hölzer and K. Schröerfor much help and expert advice inthe use of the six-circle Huberdiffractometer. We are grateful to theESRF for generous allotments ofbeam time (Projects MI-117 and HS-37), as well as to SNBL (Projects 01-02-02, 01-02-07 and 01-02-46), andto Norges Forskningsråd (NFR) forGrant 101166/431 in support of thiswork.

Fig. 3: Y-scan profiles of four triplet pairs. Estimated triplet phase, F3est, andcrystal serial no. are given in the upper profile of each pair. (a) Triplet 3 0 -4/ -24 -1/ -1 -4 5, F3CA = -48o (Cryst. #2), F3GE = -65o (Cryst. #4). (b) Triplet -2 -2 -3/ 1 4 -7/ 1 -2 10, F3CA = 105o (Cryst. #7), F3GE = -66o (Cryst. #8).

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

37

yields information about theshape and sizes of catalyticparticles adsorbed on surfaces aswell as on the interaction betweenparticle and support surfaces.Prins and his group (ETH-Z)have recently characterized HydrideTreatment Catalysts (HTC), usingEXAFS amongst other methods. Thesalient features of these catalysts aretheir high specific nickel surface area,together with their high reducibility.This uncommon combination of highnickel dispersion with low (buteffective) interaction of metal withsupport raised questions as to how theAl2O3 support stabilizes the metalparticles. Absorption spectra werecollected in the oxidic, reduced andpassivated states. Figure 1 and Figure 2show the results obtained for the oxidicstate. The Ni cations in the NiO particlesare surrounded by 6 oxygen anions as inbulk NiO; the Ni-O distance (2.04 Å)inferred is close to the distance found in

bulk NiO (2.09 Å). As Ni loading ofHTC was reduced, the coordinationnumber due to the Ni-Ni contributiondecreased dramatically. This smallcoordination number for Ni-Ni suggeststhat the NiO particles on the Al2O3support are small. From a comparisonwith various NiO particle models ofvarying size and configuration (rangingfrom Ni (111) layer to cubic NiO (100)particles) it was concluded that the NiOparticles in the oxidic catalysts werebuilt up of successive (111) layers. Thisshape of the NiO particles implies thatthey interact strongly with their Al2O3support. This was confirmed by theobservation of a Ni-Al contribution atlow Ni loading. A possible model forsuch a particle is shown in Figure 3.

The Swiss-Norwegian collaborationwas born at a workshop hosted by theESRF in Frankfurt, Germany, inFebruary 1989. On this occasion, in ahallside encounter, C. Riekel (ESRF)suggested to H.-P. Weber, Universityof Lausanne (UNIL), and F. Mo,University of Trondheim (UNIT), thatthey pool their resources and constructa bi-national beamline at the ESRF.On the Norwegian side, this idea wasquickly given form by F. Mo andD. Nicholson in a proposal thatinvolved additional commitmentfrom UNIT to basic and appliedresearch in physics, chemistry andmaterials science. The then Rector,Prof. R. Lenschow, and DirectorH. Skaar enthusiastically supportedthe proposal and submitted it forapproval to the Collegium of theUniversity. On 5 February 1990,UNIT, as the first institution,allocated 2 mill. NOK towards theconstruction of the Swiss-Norwegianbeamline.The Swiss side (next to H.-P. Weber,G. Chapuis and M. Fehlmann) was atfirst surprised by the alacrity withwhich their partners-to-be hadconverted a good yet vague idea intoa fully-fledged project. However,when the Swiss counterparts (at firstUNIL and the Swiss Federal Instituteof Technology in Zurich) realizedhow modest the beam time fraction,to which Switzerland, as a fullmember of the ESRF, was entitled,they quickly moved into action andstarted raising funds. The ensuingSwiss-Norwegian Beam LinesConsortium was approved and fullyfunded by the respective nationalresearch councils and universitiesby the autumn of 1990. Theuniversities of Oslo, Berne, Geneva,Stavanger College and the Swisspharmaceutical industry (Hoffmann-LaRoche, Novartis) joined as fullconsortial members at a later date.

SOME HISTORY

2

k / Å-1

5.0

0

2 2 2 2 2 2 2 2

k3 •

χ(k)

Fig. 1: k3-weighed EXAFS function of HTC-1.9 catalyst in the oxidic state.Obtained from the raw absorption data by subtracting background andnormalizing to one absorber atom.

Fig. 2: Fourier-transformed k3-weighed EXAFS function. This

radial distribution functionreveals the number of atoms

located at distance R fromabsorber atom (coordination

number); the distance displayedis about 0.2 - 0.5 Å shorter than

the actual distances due to aphase factor.

Al2O3(111)

O layer on the particle

Ni4-7 layer

O layer between the particleand support (originatedfrom the support)

Al vacancy

Affectedby support

Fig. 3: Possible structure ofthe NiO particles.

0R / Å

20

15

10

5

0

-5

-10

-15

-20

FT

[k3

• χ(

k)]

2 4 6

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

38

D2AM (FRENCH BEAMLINE) ON BM2:PROTEIN CRYSTALLOGRAPHY AND MATERIALS SCIENCE

The D2AM beamline has been operated up to now as a «double» beamline, dedicated to two

different communities, who exchanged instruments around every 6 weeks: on the one hand protein

crystallography using the multiwavelength anomalous method and, on the other hand, materials

science using diffuse anomalous scattering at small and wide angles.

PROTEINCRYSTALLOGRAPHY

Numerous structures have beenquickly solved using data collected withthe protein crystallography instrumentor with the help of these data. The beamintensity is very high, D2AM being oneof the most intense European beamlinesfor this application apart from the ESRFundulators, ID2, ID9 or ID14. But sincethe beam comes from a bending magnet,high intensity is obtained from beamfocusing. Therefore the beamline is notvery well suited for experiments withcrystals having unit cell parametersabove 300 Å, as they require very lowdivergence. A MSCTM cryogenic systemis attached to the goniometer; it allowsone to work at temperatures of – 170 °Con crystals obtained by flash in situcooling or on pre-cooled crystals: veryfew protein crystals can be irradiated atroom temperature without damage. Thex-ray image intensifier with CCDdetector gives high quality data but theexposure time cannot be increased over100 s; typical exposure times arebetween 5 and 60 s according to crystalsize and quality with a dead time of 15to 20 s between images.

New and original research carried outon the beamline is concerned with thestudy of the structure of the anomalousdiffraction near an absorption edge withprotein to use numerous data taken atwavelengths near an absorption edge.This implies direct relative scaling of allthe data sets, which gives as asecondary result a measurement of thecrystal fluorescence. This should be avery powerful phasing method, but itrequires a very large amount of data,as at least 12 sets at differentwavelengths have to be collected [1, 2].The beamline has also been used formosaicity studies; a specific softwarehas been developed; these measurementshave been used to analyse the advantageof growing crystals in microgravity [3].

MATERIALS SCIENCEThe materials science group extends

its activity over a wider domain. One partis concerned by small-angle scattering,the other with crystallographic activities.The instruments were built not only foranomalous scattering but also to improvethe resolution and the dynamic rangeavailable on laboratory instruments. Thisgoal has been reached. The following

report does not represent the wholeactivity but gives some light on specificthemes. D2AM initiated DAFSspectroscopy at the ESRF and hastransformed it into a techniqueaccessible to outside users: appliedto mutilayers, this site-resolvedspectroscopy allows the measure of theinter-diffusion between layers. Phasondisorder has been studied in AlPdMnquasicrystals, diffuse scattering wasrecorded over more than 8 orders ofmagnitude; the results can be wellreproduced by the elasticity theory ofquasicrystal, considering only longwavelength phason fluctuation [4-6].The mechanical properties of superalloysare related to a large volume fraction ofan ordered phase that is coherent with thedisordered fcc matrix. Their processesare linked to the kinetics of systems farfrom equilibrium. The NiCrAl systemappears as a model alloy: the first statesafter quench from a homogeneous alloyhave been studied both by small-angleand wide-angle anomalous scattering. Inthis ternary alloy, the first overallordering is a high Cr content located onAl sites; during growth of precipitates,Cr is replaced by Al. Meanwhile, thecomposition of the matrix is fairly

From left to right:P. Carpentier,M. Pirrocchi,S. Arnaud, J.F. Berar,B. Caillot, J.P. Simonand M. Roth.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

39

constant. This behavior should beinterpreted as a nucleation of thedisordered phase after preliminaryordering of the whole, and gives anexperimental answer to the question:«what occurs first: ordering or phaseseparation?». The relation to models is,however, less straightforward: theore-tical models have been confined tobinary systems.Studies of dynamics around a phasetransition have also been performed,showing how alloys come back to theirthermodynamic equilibrium; this kindof experiment requires both a very highresolution and a high intensity. Anexample is given hereafter.

M. Roth, J.F. Berarand J.P. Simon

The building of the French CRGbeamlines, D2AM and IF, startedsimultaneously in 1990 with thefirst ESRF beamlines. They wereestablished with the participation ofthe Rhône-Alpes region, the IsèreDepartment and the «Ministry ofNational Education and ScientifcResearch». Both the CEA and theCNRS are in charge of these beamlines,providing their financial support andmanpower for construction andexploitation.The D2AM beamline is the resultof two independent initiatives thatconverged in their technicalrequirements. M. Roth (LCCP/IBS)required an instrument dedicated to thecollection of three-dimensional atomicstructure data of protein crystalsusing the multiwavelength anomalousdiffraction method: this impliesfrequent wavelength changes as well ashigh intensity on the sample. J.P. Simon(LTPCM/INPG-CNRS) and a group ofmaterials science laboratories fromGrenoble and Orsay were interestedin investigating the diffuse anomalousscattering at small and wide angles.They needed a more versatileinstrument but with demanding opticsrequirements: both the harmonic purityand the signal-to-noise ratio had to beimproved. This explains why twocommunities have shared D2AM.However, the situation will beimproved in 1998 when the new FrenchFIP-CRG beamline will be completed.

SOME HISTORY

REFERENCES

[1] G. Boissy et al, ESRF Newsletter 28(1997), p. 18-20.[2] E. Fanchon et al, ESRF Newsletter 24(1995), p. 6-7.[3] J.L. Ferrer et al, ESRF Newsletter 26(1996), p. 27-29.[4] H. Renevier et al, ESRF Newsletter 23(1995), p. 7.[5] D. Sayers et al, ESRF Newsletter 27(1997), p. 18-19.[6] ESRF-Highlights 96/97, p. 38-39.

THE HIGH TEMPERATUREPROJECT

C. LANDRON1,2, L. HENNET1,J.P. COUTURES1, M GAILHANOU2,M. GRAMOND3 AND J.F. BERAR4

1 CENTRE DE RECHERCHES SUR LES MATÉRIAUX À

HAUTE TEMPÉRATURE, ORLÉANS (FRANCE)2 LURE, ORSAY (FRANCE)3 ECOLE CENTRALE DE PARIS, CHATENAY (FRANCE)4 LABO. DE CRISTALLOGRAPHIE, GRENOBLE (FRANCE)

First containerless characterizationusing combined XAFS and XRD onrefractory oxides at very hightemperature are reported. We haveperformed in situ experiments at theD2AM beamline on solid and liquidrefractory oxides by using a laser heatingsystem and aerodynamic levitation. Thispaper describes the experimental deviceand presents some of the first resultsrecorded during dedicated shifts.

A two crystal monochromator– Si(111) in our experiments – provides amonochromatic x-ray beam in the 5-25 keV energy range. The beam isfocused to a size of 0.5 x 0.5 mm2 by bentmirrors and by the second crystal of themonochromator. The wavelength can becontinuously adjusted as needed inDAFS experiments, which makes itpossible to record absorption anddiffraction data in exactly the samesample conditions. We have developed aset-up based on laser heating andaerodynamic levitation in order to gain aninsight of materials at very hightemperature. The aim of the experiment isto investigate the structural properties,especially short and long-range order, inpartially disordered specimens. Theanalysis chamber, which has beendesigned for working on the 7-cradlegoniometer of the D2AM station,consists in an isolated cell performing insitu experiments combining x-rayabsorption and diffraction. Up to now,three teams had used levitation

(electromagnetic, aerodynamic oracoustic) associated with suitable heating(electromagnetic, laser ir-radiation) [1-3],but no one has combined two or moreamong these x-ray spectroscopies. Ourfirst measurements at the ESRF havebeen carried out in November 1997. Westudied the structure of some solid andliquid refractory oxides at temperaturesup to 2750 °C: ZrO2, melting temperatureTm = 2715 °C; Y2O3, Tm = 2440 °C;YAG, yttrium aluminum garnet, Tm =1945 °C; Gd2O3,, Tm = 2440 °C; Ho2O3,Tm= 2420 °C and Er2O3, Tm = 2420 °C.

The levitation of the spherical sampleis performed with a gas jet (argon orhelium) to steadily bear a sphericalsample in an energy well. A convergent-divergent diffuser creates a stable vortexring that traps the sample. This levitatorallows the sample to float in a centralposition independently of any contact. Asimilar system has been used for shapingthe spherical sample from compactpowders that have been used in x-raycharacterization experiments. Theheating is obtained by irradiation of thesample by a continuous wave (cw) 100 WCO2 laser (SYNRAD). For securityreasons, the invisible CO2 laser beamwas displayed by a red He-Ne laser beam.Our device has been configured in orderto provide a large open solid angle for thephotodiodes used as detectors for XAFSexperiments in fluorescence mode.

X-ray absorption spectra have beenrecorded in the fluorescence modewith large area silicon photodiodes(Hamamatsu). The position of thedetectors has been adjusted close to thesample in order to achieve an optimumfluorescence signal emitted by thesample. The photodiode supports werewater-cooled. Beryllium windows havebeen used to transmit x-rays and toabsorb visible light. Figure 1 illustrates

5k, Å-1

k2-χ

(k),

A.U

.

LIQUID YAG

10

0.5

0.0

-0.5

Fig. 1: k2χ(k) EXAFS oscillations atthe yttrium K-edge for a liquid dropof YAG (yttrium aluminum garnet).Oscillations above k = 8 Å-1 are thesignature of the Y-Y bonds.

the k2χ(κ) function of the yttrium K-edge recorded at 2200 °C above meltingof an yttrium aluminum garnet (YAG)sample. The signal-to-noise ratio seemsto be sufficiently high to exhibit theEXAFS oscillations above k = 8 Å-1

characteristic of the yttrium-yttriumbond.

In order to perform time-resolveddiffraction measurements during

heating and cooling, we used a PositionSensitive Detector (PSD) developed byJ.-F. Berar [4]. A gas (argon) detector isused with an angular aperture of 30°(angular resolution in 2θ of 0.1°)located at a distance of 130 mm of thespherical sample. The measurement ofthe position of a photon results from thedetermination of the delay time betweentwo signals emitted by the incoming

photon. The 50 diffraction patternspresented in Figure 2 have beenrecorded during the cooling time of2.5 s (50 ms per diagram) of aliquid drop of Ho2O3, showing thesolidification of the drop. From thediagram, we suggest the followingphase transitions: the first, towards amixing of high-temperature cubic andhexagonal phases, the second towards amixing of low-temperature cubic andhexagonal phases and the last, towards alow-temperature cubic phase.

We have also recorded diffusionspectra on various liquid oxides with ascintillator coupled to a photomultiplier.The spectra were recorded on anangular range of 120° givingcomplementary information to the pairfunction data calculated by EXAFSanalysis. The interest of measurementof x-ray diffusion oscillations is due tothe reliability of the informationobtained at low k values. We can safely conclude through thesefirst encouraging results that gaslevitation associated to laser heating is apotential tool for contactless processingof refractory materials with particularproperties controlled by in situsynchrotron analysis at high temperature.The real-time characterization bydiffraction allows the follow-up ofphase transition phenomena as rapid as50 ms.

REFERENCES

[1] S. Ansell, S. Krishman, J.K. Weber,J.J. Felten, P.C. Nordine, M.A. Beno,D.L. Price and M.L. Saboungi, 1997, Phys.Rev. Lett., 78: p. 464-467.[2] G. Jacob, I. Egry, K. Maier, D. Platzek,J. Reske and R. Frahm, Rev. Sci. Instrum. 67,p. 3683-3687. (1996) [3] C. Landron, X. Launay, J.C. Rifflet, Y. Auger,D. Ruffier, J.P. Coutures, M. Lemonier,M. Gailhanou, M. Bessière, D. Bazin andH. Dexpert, 1997, Nucl. Instr. and Meth. B 124,p. 627-632.[4] J.F. Berar, M. Lemmonier, F. Bartol,M. Gramond. and J.Chevreul, 1993, Nucl.Instr. And Meth. B, 82, p. 146-150.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

40

ACKNOWLEDGEMENTS

The contribution of D. Ruffier andY. Auger in the construction of theanalysis cell is highly acknowledged.We would like to gratefully thank theESRF staff. This work was financiallysupported by a contract «XI ème planEtat / Région-Centre».

20

2 θ

Inte

nsit

y

Ho2O2

100

0

30 40 50 60

Fig. 2: 50 successivereal-time diffractionpatterns recorded duringthe cooling of a drop ofliquid Ho2O3. The recordtime is 50 ms perdiagram. L = liquid, HTC= high-temperature cubicphase, LTC = low-temperature cubic phase,H = hexagonal phase.

Atoms, molecules and more complexstructures are studied almost uniquelyby the long-range electromagneticinteractions, while the much smallernuclei and elementary particles havebeen investigated mostly by the short-range strong and weak interactions. Infact, if the strong, hadronic interactionyields the largest cross-sectionsand therefore the highest event rate,photon-induced reactions have severaladvantages, and high-energy gammabeams can be good probes of nuclei.

The first polarised and taggedgamma-ray beam was obtained by thebackward Compton scattering of laserlight against high-energy electronscirculating in a storage ring (Adone, inFrascati). After this successful start,other similar beams entered intooperation at Novosibirsk andBrookhaven. When the ESRF wasoriginally proposed, it was evident thatits high energy and low emittancewould make it the best machine toproduce such a beam and therefore aproposal was immediatelysubmitted and included in allESRF presentations. It wasnamed Graal.

The Graal experiment isnow in full operation. Itproduces a gamma-ray beamwith a maximum energy of1470 MeV and an intensity ofabout 106 photons per second. Ithas a 98% linear polarisation atthe maximum energy and aresolution of 16 MeV fwhm.The main parts devoted to theproduction and monitoring ofthe beam are:• a laser electron interactionregion situated in one of theshort straight sections of thestorage ring;• a tagging system, to measurethe position and therefore theenergy of the electrons whichhave interacted with the laser

photons. It is located at the exit of thestorage ring dipole which follows theinteraction region;• a vacuum system to connect themachine vacuum pipe to the cabin wherethe laser is located;• a laser cabin with the optical benchwhich supports the laser and its optics;• an adjustable lead collimator, aclearing magnet and a second, fixed,lead collimator to collimate the gamma-ray beam and clear it of the backgroundof undesired photons and electrons;• a liquid hydrogen (or deuterium) target;• three intensity, energy and positionmonitors for the gamma-ray beam.The detector consists of:• a BGO crystal ball made of 480 BGOcrystals for the detection of high-energygamma-rays and medium-energyprotons; it covers a polar angle between25° and 155°;• two cylindrical wire chambers locatedaround the target, inside the crystal ball,followed by a barrel made of 32 plasticscintillators. The wire chambers and the

barrel are used to measure the angles andthe ionization of the charged particlesentering the BGO;• two plane wire chambers and threewalls of plastic scintillators to detectparticles emitted in the forwarddirection, at θ < 25°.

This very complicated apparatusrequires various expertises: storage ringphysics, laser optics, gamma-ray beamhandling, cryogenic targets, organic andinorganic scintillator detectors and wirechambers, plus fast electronic, dataacquisition electronics, data handling,etc. This expertise is provided by alarge international collaboration whichincludes about 50 scientists from Italy,France, Russia and the USA.

The beam polarization asymmetrieshave been measured for the photo-production of π°, π+, 2 π° and η. A moredetailed report on these measurementswill be given in a future issue of theESRF Newsletter.

C. Schaerfand D. Rebreyend

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

41

GRAAL ON BM7: GAMMA-RAY BEAM

GRAAL does not exploit synchrotron radiation but utilizes gamma-rays

produced by Compton backscattering of photons from the circulating electrons,

for studies in nuclear and particle physics.

As such, it is not a typical CRG beamline, for example no

beam time is made available to general ESRF users.

The BGO crystal ball surrounding the cryogenic target on Graal.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

42

T

GILDA (ITALIAN BEAMLINE) ON BM8

The Italian CRG beamline GILDA is financed by three major Italian public research institutes

(Consiglio Nazionale delle Ricerche CNR, Istituto Nazionale per la Fisica della Materia INFM and

Istituto Nazionale di Fisica Nucleare INFN). It has been operational since September 1994 and is

mainly dedicated to the investigation of local structure. To this purpose, x-ray absorption spectroscopy

as well as x-ray diffraction are used on the beamline in the energy range 5-50 keV.

he beamline GILDA consistsof four experimental hutches. The firstexperimental hutch, in the 1:3 focalconfiguration which ensures themaximum flux, is dedicated to X-rayAbsorption Spectroscopy (XAS). Thewide energy range of the beamlinemakes GILDA well-suited for XASinvestigations on heavy elements withthe possibility to access the Kabsorption edges from Ti (Z = 22) to Eu(Z = 63). Experiments are carried out intransmission, fluorescence and totalelectron yield modes. For the study ofhighly diluted samples a 7-elementhigh-purity Ge detector is currentlyused (current limit sensitivity: 1014

at/cm2). Surface experiments in totalreflection mode (ReflEXAFS) areperformed in a dedicated experimentalchamber.

The second hutch, in 1:1 focalgeometry, is dedicated to x-rayscattering and diffraction. It is equippedwith a two-circle diffractometer with anangular step of 0.28 arcsec and areproducibility of 2 arcsec. Crystalanalyzers, solid-state detectors andscintillators are used to performanomalous scattering on amorphousmaterials, and powder diffraction withan instrumental angular resolutionlower than 0.01 degree. The installationof an Image Plate detector is foreseen.Anomalous x-ray reflectivity spectrahave also been collected. Finally,Diffraction Anomalous Fine Structure(DAFS) spectra have been successfullyrecorded: soon also this technique willbecome available for users.

A third experimental hutch at theend of the beamline is available forusers who wish to install their ownapparatus. An ultra-high vacuumchamber fully equipped for surfacepreparation and characterization isavailable for EXAFS investigations intotal and partial electron yield mode.

SCIENTIFIC ACTIVITY

The scientific activity of GILDAmainly deals with the study of localstructure around atoms present in traceamounts. XAS is a well-suited techniquefor this: among the experiments performedwe recall here the investigations onInAsP/InP and InGaAs/InP superlattices,impurities in semiconductors (Er:Si, Ga:a-Ge, As:Si, Co:Si), impurities (Ag, Cu, Er,Eu, Pr) in various kinds of glasses,metallic active sites in proteins. Catalyticsystems have been investigated in cellsfor in situ treatments. The 1997 ESRF-Highlights booklet reported the GILDAstudies performed on the metalinsulator phase transition in colossalmagnetoresistance manganites as wellas epitaxial overlayer of InGaAs/InP.

X-ray scattering experiments onamorphous and polycrystalline sampleshave been performed in theconventional and anomalous modes toobtain a description of the local andmedium range order of the wholematerial and around a particularcomponent. Experiments carried out inthis field provided the local orderaround Eu and Sr in Eu-doped Sr-

metaphosphate glasses, around themetallic components of FeNiBamorphous alloys and around Zr in Zr-based ceramics. Other kind ofexperiments performed were anomalouspowder diffraction on nanocrystallinemetallic clusters dispersed on a pumicesubstrate, anomalous x-ray reflectivityin InAsP/InP superlattices and DAFS onZnAl2O4 and ZnFe2O4 spinels.

LOCAL ATOMIC ORDERAROUND ER IN ER-BASEDLUMINESCENT MATERIALS

Er doped crystalline Si, exhibitingoptical luminescence at 1.54 µm thatcorresponds to the absorption minimumof commercial optical fibers, represent apromising material for optoelectronicalapplications. The luminescence arisesonly from Er+3 ions bound to O atoms(usually present in Si as impurities) andthe effect is absent in Er-doped high-purity Si. A new way to obtain a highconcentration of Er+3 active centersconsists in performing a doubleimplantation of Er and O ions on acrystalline Si substrate. The as-implanted material undergoes thenseveral annealing treatments torecrystallize the implanted region.Unexpectedly, the as-implantedmaterial exhibited a barely detectableluminescence, strongly increased by theannealing process. To understand thisphenomenon, an investigation of thelocal atomic order around the Er+3 ions

1 CNR ROMA

2 INFM-GENOVA

3 UNIVERSITY OF PARMA

4 INFN FRASCATI

5 UNIVERSITY OF TRENTO

6 UNIVERSITY OF ROMA «TOR VERGATA»7 UNIVERSITY OF CAGLIARI

8 CNR TRENTO

9 UNIVERSITY OF ROMA «ROMA TRE»

F. D'ACAPITO1, S. COLONNA2, S. PASCARELLI2, G. ANTONIOLI3, A. BALERNA4,A. BAZZINI3, F. BOSCHERINI4, F. CAMPOLUNGO4, G. CHINI5, G. DALBA5,

G. I. DAVOLI6, P. FORNASINI5, R. GRAZIOLA4, G. LICHERI7,C. MENEGHINI2, F. ROCCA8,L. SANGIORGIO4, V. SCIARRA4,

V. TULLIO4 AND S. MOBILIO9

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

43

was performed using the Extended X-ray Absorption Fine Structure (EXAFS)technique at the Er LIII edge. Thetypical amount of Er ions in thesesamples was around 2*1015 Er/cm2

(roughly 2 ML). In the as-implantedmaterial (process A - Figure 1a) theposition of the first FT peakcorresponds to an Er-Si coordinationwith no evidence of Er-O bonds. Er andO do not react, probably due to theextremely reduced diffusion constant ofEr in the Si network, and thecorrespondent luminescence is very low(Figure 1b). The spectra of the annealedsamples (B and C) on the other handshow a peak corresponding to an Er-Obond and the luminescent yields aremuch stronger. Upon annealing, thecrystal-amorphous boundary movestowards the surface, creating a region ofhigh atomic mobility where theimplanted species can react togetherand give rise to the active centers.

LOCAL STRUCTUREOF GA-DOPED A-GE:H

THIN FILMSOne of the major issues in the

physics of amorphous semiconductorsis the understanding of microscopicmechanism of substitutional doping inhydrogenated amorphous Si and Gethin films doped with group III and Velements. In this work hydrogenated

amorphous Ge thin films doped withdifferent Ga concentrations have beeninvestigated by EXAFS spectroscopy.The films were prepared by rfsputtering from a c-Ge target, dopantswere added using a co-sputteringtechnique. The typical thickness of the

films was around 3 µm, with a Gaamount ranging between 5 x 1014 and1.3 x 1017 at/cm2. Data were collectedin fluorescence mode. From thequantitative analysis, the coordinationnumber of Ga at the lowerconcentration was found to be 4, in

R (Å)0

Fou

rier

tra

nsfo

rm

1 2 3 4 5 6 1.52 1.54 1.56 1.58 1.60

Wavelength (µm)

PL

inte

nsit

yEr2O3

C

B

A

ErSi2

Er2O3 ( 200)

C

B

A (x5)

Fig 1: a) Fouriertransforms of the EXAFSspectra compared withmodel compounds for theEr-Si bond (ErSi2 crystal,peak at ~2.5 Å) and the Er-O bond (Er2O3 crystal peakat ~2 Å). b) High resolutionphotoluminescence spectraat 15 K with a 200 mWlaser pump. Capital lettersindicate the thermalprocesses: A = 450°C-30min, B = A+620 °C-180min, C = B+900 °C-30s.

Fig 2: The upper panelreports the Total

Correlation Functiong(r) for

Eu0.1Sr0.9(PO3)2.1obtained at 25 keV.

The middle panel showsthe Differential

Correlation Function∆g(r) obtained at the SrK edge and in the lower

pannel is reported theDifferential CorrelationFunction ∆g(r) obtained

at the Eu K edge.

R (Å)

Arb

. uni

t0 2 4 6 8

g(r)E = 25 KeV

∆g(r)Sr

E1 = 15.8 KeV

E2 = 16.097 KeV

∆g(r)Eu

E1 = 46 KeV

E2 = 48.480 KeV

a b

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

44

agreement with the hypothesis ofsubstitutional site of the dopant inthe Ge tetrahedral network. A sharpdecrease in the mean coordinationnumber from 4 to 2.8 occurs when thedopant concentration rises from4.5 x 1015 to 4.5 x 1016 at/cm2

suggesting a Ga site modification athigh concentration.

SR AND EU DOPED META-PHOSOPHATE GLASSES

Anomalous Wide Angle X-rayScattering (AWAXS) and extended x-ray absorption fine structure (EXAFS)have been used to investigate thestructure of Eu doped Sr metaphosphateglass (Figure 2). In order to investigateboth the local structure around Eu andthe modifications of metaphosphatenetwork induced by doping twoglass samples, a-Sr(PO3)2 and a-Eu0.1Sr0.9(PO3)2.1, were studied at theSr (16.107 keV) and Eu (48.517 keV) Kedges. The resulting model of the hostSr-metaphosphate matrix is consistentwith a description of the structure

in terms of Sr2+ ions interposedbetween PO4 chains. All distancesand coordination numbers related toa distribution of PO4 tetraedra,interconnected via two bridgingoxygens, fall within the expected results.However, the location of strontium ismore problematic, and the coordinationnumber NSr-O ~ 5 suggests an irregulardistribution of its first coordinationshell or an equal number of four-foldand six-fold coordinated Sr atoms.The data obtained on both samplesdemonstrated that the host metaphosphatematrix is unchanged by doping. Thedetermination of the local structurearound Eu by AWAXS at the Eu K edgehas been a special issue of thisexperiment as it is the first attempt toperform an AWAXS experiment at suchenergies. The choice of one of the threeEu-L edges is also possible to detect theEu environment, but several problemsarise in this case: for the low energy,the reciprocal-space extension of thedata is not enough to allow a goodresolution of the Eu coordination shells.Moreover, the three L absorption edgesare very close each other, making the

measurements more difficult for thelower contrast between the scatteringfactors of Eu and high fluorescencesemission. AWAXS results demonstratedan asymmetric environment aroundEu with 7.5 O at 2.32Å and 1.5 at2.62 Å.

REFERENCES

- A. Terrasi, G. Franzò, S. Coffa, F. Priolo,F. d’Acapito, S. Mobilio App. Phys. Lett. 70(1997), 1712.- M. Bionducci, C. Meneghini, G. Navarra,G. Licheri, A. Balerna, S. Mobilio, MRS Symp.Proc. 437, 161-167 (1996).- A. Balerna, M. Bionducci, A. Falqui, G. Licheri,C. Meneghini, G. Navarra, M. Bettinelli acceptedby J. Non Cryst. Solids (1997).

ACKNOWLEDGEMENTS

We are grateful to F. Berti, A. Gasparoli,D. Gasperetti, P. Gennara, P. Recchia,A. Rossi and G. Silvestrin fortechnical support in the constructionand commissioning of the x-rayabsorption set-up.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

45

To fulfil its purpose, besides the opticshutch and a small beamline workshop,ROBL has two experimental stations:the radiochemistry hutch (RCH) and thematerials research hutch (MRH). TheRCH is built as a radiochemicallaboratory of type B to performexperiments with radioactive sampleshaving an activity up to 185 MBq (5mCi). Such a dedicated installation isunique in the world at a synchrotronradiation source.

THE RADIOCHEMISTRYHUTCH (RCH)

The Institute of Radiochemistry isinterested in research for riskassessment and remediation ofradionuclide contaminations related touranium ore mining and milling, ofenvironmental contamination resultingfrom world-wide nuclear weaponproduction activities and nuclear

accidents. Furthermore, it is interestedin research for risk assessment ofpotential nuclear waste repositories.A basic molecular-level understandingis required to quantitatively describe themechanisms of radionuclide transportin the environment. XAS is a powerfultechnique to study speciation andcomplexation of radionuclides insolutions, adsorption processes at thesolid-water interface of radionuclidesolutions with geological and biologicalmaterials.

The radiochemistry hutch (RCH) isdesigned to handle radioactive samples(solid or liquid) with an activity of up to185 MBq (5 mCi). The samples containthe actinides from Th to Am as well asRa, Po and Tc.

The main experimental method willbe x-ray absorption spectroscopy(XANES and EXAFS) in both thetransmission- and the fluorescencemode (for diluted samples). A closedcycle He cryostat allows for studies in

the temperature range of 4-300 K.Quick EXAFS for time-dependentstudies and microfocus x-ray absorptionspectroscopy for spatially-resolvedstudies are planned. Different detectionsystems (ion chambers of differentlengths, Ge solid state fluorescencedetectors, fluorescent x-ray ion chamberdetectors) will be available.

All radioactive samples will bedelivered, stored and handled accordingto the safety regulations agreed upon withthe ESRF. Only specially trained staffwill be allowed to operate thisexperimental station. However, ourexpertise and help can be made availableto collaborators.

W. MATZ, N. SCHELL, H. FUNKE AND G. BERNHARD

ROBL (GERMAN BEAMLINE) ON BM20:STRUCTURAL AND RADIOCHEMICAL INVESTIGATIONS

The beamline is located at the bending magnet BM20.

The two experimental stations, operating alternatively, are designed for radiochemical investigations

with x-ray absorption spectroscopy on the one hand and structural investigations of materials by x-ray

diffraction and reflectivity on the other hand.

SOME HISTORYIn the beginning of the year 1995the Forschungszentrum Rossendorf(FZR) near Dresden, Saxony, signed acontract with the ESRF to establishthe German CRG ROBL in order toenhance the research capabilitiesof the biggest research institution ofthe new Länder in reunified Germany.Building and running of ROBL willbe carried out by close co-operationof the Institute of Radiochemistry andthe Institute of Ion Beam Physics andMaterials Research as well as theCentral Department of Research andInformation Technology of the FZR.Since May 1996 a small permanentstaff is residing at the ESRF tosupervise and contribute to thebuilding progress. The constructionphase in the ESRF experimental hallstarted in May 1996.Tests and preliminary experiments byFZR groups started early in 1998. Thestart of scheduled user operation ofROBL is planned for September 1998.

Staff situated in Rossendorf from left to right: S. Dienel, D. Pröhl, W. Oehme,R. Schlenk, W. Matz, P. Reichel, M. Betzl, M.A. Denecke, V. Brendler, G. Hüttig,F. Eichhorn, G. Bernhard, T. Reich, J. Claußner.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

46

THE MATERIALSRESEARCH HUTCH (MRH)

The Institute of Ion Beam Physicsand Materials Research will use ROBLwithin its scientific program for theidentification and characterization ofthe modifications of surfaces, interfacesand near surface layers including phaseformation produced by ion beamtechniques. Research topics of theinstitute are hard coatings, nanoclusters,biocompatible materials and advancedsemiconductors.

In collaboration with universitiesand other research institutes from theregion structural investigations ofmelts, amorphous solids and phases inmetal and semiconductor nanometer-multilayers as well as the analysis of thereal structure of single crystallites ortexture are planned. But also other usersare invited to perform experiments inMRH in collaboration with the FZR. The basic equipment in MRH willconsist of a high-precision 6-circle Huber goniometer, severaldetectors (scintillator, energy sensitivephotodiode, two-dimensional PSD of

CCD-type) and a high-temperaturechamber up to 2000 °C.

REFERENCES

[1] More information is available underhttp://www.fz-rossendorf.de. Postal address:Forschungszentrum Rossendorf, P.O. Box510119, D – 01314 Dresden.[2] ESRF Beamline Handbook; also athttp://esrf.fr/exp_facilities/BLHB.htm.[3] Contact ESRF residents of ROBL atrobl@esrf.fr.

ACKNOWLEDGEMENTS

We would like to thank the ESRF fortheir support in beamline planningand installation matters, and the ILLfor agreeing to the temporary storageof future radioactive samples.

SPLINE (SPANISH BEAMLINE) ON BM25At its July 1997 meeting, the ESRF Council has approved the construction of the Spanish CRG

beamline (Spline) financed by the CICYT (Comisión Interministerial de Ciencias y Tecnología).

The main goal of the Spanish CRG x-ray beamline is to satisfy the needs of the Spanish scientific

community and give it access to a third generation synchrotron radiation facility to perform x-ray

absorption and diffraction experiments in a broad energy range.

BRANCH A:• High-resolution powder

diffraction (HRPD)including anomalousdispersion.

• X-ray absorption spectroscopy (XAS) and x-ray standing waves (XSW).

BRANCH B:• Macromolecular crystallography

including MAD• Single crystal diffraction

and diffraction from interfaces• X-ray diffraction/scattering camera

for non-crystalline specimens.

The hutches are due to beconstructed in January 1999. It isintended that the first station willbe ready for user operation inJanuary 2001. Full operation willstart in the second half of 2002.

More in ESRF Beamline Handbook; alsoat http://esrf.fr/exp_facilities/BLHB.htm

Contact: castro@esrf.fr

The Spanish beamline will be split into two lines:one branch will be allocated on the soft edge (A)

and the other on the hard edge (B) of the BM25 bending magnet.

Local staff of ROBL at the ESRFwith project leader W. Matz (left):N. Schell (Materials Research),H. Funke (Radiochemistry)and U. Strauch (technical service).

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

47

T

A

M. OVERSLUIZEN, P. GOEDTKINDT, W. BRAS, D. DETOLLENAERE, R. VAN TOL AND W. DE JEU

WITH SUPPORT FROM

I. CERJAK, D. GLASTRA VAN LOON, P. LASSING, B. MUNNEKE, J. DERKS, M. KONIJNENBURG,M. BORSBOOM, H. ALBERDA, R. BAKKER, F. UDO

AND SEVERAL OTHER TECHNICIANS, ENGINEERS AND SCIENTISTS

DUBBLE (DUTCH-BELGIAN BEAMLINE) ON BM26

lthough neither Belgium nor theNetherlands have a national facility forthe use of synchrotron radiation,researchers from these countries havebeen using synchrotron radiation as atool for a long time. Lure, Desy andDaresbury in particular have regularlybeen hosts for scientists from thesecountries. The Netherlands ResearchCouncil (NWO) has been an activepromoter of the use of synchrotronradiation notably through its formalinvolvement in Daresbury where twobeamlines were constructed and severalother Dutch/English projects receivedfinancial support. For over ten yearsNWO-funded scientists were part of thestaff in Daresbury. When the CRGprogram at the ESRF was approvedattention shifted towards Grenoble. Afterextensive consultations with possibleusers a consortium was formedconsisting of NWO, the universities ofUtrecht and Amsterdam and AMOLFfrom the Netherlands, and NFWO andthe universities of Leuven, Antwerp,Gent and Brussels in Belgium. Thedecision was made to build twobeamlines, which after the decision fromthe ESRF to support the development of

the wide 9 mrad front-ends, will beequipped with full focusing optics onboth beamlines. The experimentaltechniques that will be implemented areinterface diffraction, powder diffractionEXAFS and time-resolved simultaneousSAXS/WAXS. This reflects the need of alarge part of the Dutch and Belgian usercommunity. The DUBBLE project willconsiderably enhance the number ofshifts available to Dutch and Belgianusers at the ESRF.

The design and construction is donemainly at the Amsterdam-basedinstitutes AMOLF and NIKHEF.NIKHEF is the Dutch high-energyphysics institute which had already beeninvolved in the construction ofsynchrotron radiation beamlines. A teamof four scientists and two technicians isresponsible for the building andcommissioning of the beamlines butthere is generous support from AMOLFand NIKHEF so that the expertise andeffort from several engineeringdisciplines is available. The collabo-ration between the high-energy physicsand synchrotron radiation experts isquite unique, especially in thecollaboration on position-sensitive

detectors. Much is expected of this.Besides the optics hutch there are

two experimental hutches. The first willhouse the EXAFS and powderdiffraction equipment, and the secondthe interface diffraction andSAXS/WAXS. The latter hutch is ratherlarge in order to house the long opticalbench for SAXS and to be able to mountlarge sample environments used forinstance in polymer processing but alsoto house the large six-circle interfacediffractometer.

The project is, after some slight delay,in an advanced state and it is expectedthat the first technique, time-resolvedSAXS/WAXS, will become operationalin the first half of 1998, followed byinterface diffraction shortly afterwards.Later in 1998 the construction of theEXAFS and powder diffraction stationswill start. This will be completelyfinished in 1999. At the moment two staffmembers are permanently based inGrenoble. This number will increase tofive in 1998 when more equipment,assembled in Amsterdam, is ready to bemounted on the beamlines. It is alsoforeseen that more universities will jointhe consortium in the coming years.

his project was conceived back in1990 when M. Cooper, B. Stirling andG. Stirling (no relation) were emptying abottle of wine at a Daresbury Users’Meeting. Three years were spent inrefining and revising an application to theUK Science Research Council (EPSRC);the budget grew by a factor of five fromour first naive estimate. The proposal waseventually accepted by EPSRC and fundsallocated for construction of the beamlinein 1994. It was not until January 1996that the first hutch was erected on thefloor in front of dipole D28 and the

construction began in earnest. A grant torun it was thankfully forthcoming fromthe EPSRC earlier this year and on 19September the transition to the newfunding was marked by further emptyingof wine bottles. As is usual in thesecircumstances, the «opened» beamline isstill being commissioned and we expectto take our first really independent usersin Spring 1998.

The acronym XMaS (X-ray MagneticScattering) was invented by B. Forsyth,but if you think it is corny take a momentto savour those of some of our

continental European colleagues (haveDubble never heard of double Dutch?).Our objective is to focus a few areas ofphysics that can peacefully co-exist on asingle beamline. It is designed primarilyfor single-crystal diffraction studies,especially magnetic diffraction studies,but also other forms of high-resolutionsingle crystal diffraction, white beamstudies and powder diffraction will bepossible. The UK scientific communitywaiting to use it is big enough to justifythe expense but small enough to ensurethat users can expect to win sufficient

XMAS (UK BEAMLINE) ON BM28:X-RAY MAGNETIC AND HIGH-RESOLUTION SCATTERING

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

48

8.00

Energy (keV)

HolmiumQ = (0 0 2 + τ)

300000

200000

100000

08.02 8.04 8.06 8.08 8.10

Pea

k In

tens

ity

(cts

/s/2

00m

A)

3.70

Energy (keV)

UAs0.78 Se0.22Q = (0 0 2 + τ)

1500

1000

500

0

Pea

k In

tens

ity

(cts

/s/2

00m

A)

3.71 3.72 3.73 3.74 3.75

time to sustain their research programs.The beamline has some features that aredifferent from other dipole lines at theESRF. First, it is aligned on the soft endof the dipole in order to maximize theflux that we get at 3-4 keV. This isspecifically in response to our interest inthe magnetic structure of the actinidesand transuranic materials whosecompounds have fascinating magneticproperties: their M-edges are at these lowenergies.

The beamline is also a white beamstation because the only viable methoddeveloped to date for measuring themagnetic form factors of ferromagnets isbased on the use of white beam. In fact aspart of the commissioning exercise wehave characterized the polarization of thesource in a white beam diffraction studyof the magnetic form factor of nickel. Itis the angular distribution of the electrontrajectories at the source point thatgoverns the beam polarization. Our studyshowed that the linear polarization, PL,of the beam from the dipole source D28is 99.85 ± 0.01% at 10 keV; thecorresponding figures at 5 and 15 keVare 99.88 and 99.83, respectively. The10 keV result, as a function of the angleout of the orbital plane at which thesource is viewed, is shown in Figure 1together with the circular polarization,PC, and the factor fP = PC/(1-PL), which«magnifies» the ratio of magnetic tocharge scattering in the white beammethod (see for example, Collins et al,Phil. Mag. B65, 37 1992). The highdegree of polarization helped us tomeasure magnetic form factors out to the(18,0,0) reflection (Laundy et al: in printJournal of Synchrotron Radiation).

The subjects identified for study bythe CRG’s user group (there are over adozen groups in the UK who plan to usethe XMaS facility) fall into two broadcategories. First, magnetic materials: themagnetic structure of lanthanide andactinide metals, compounds and multi-layers using resonant exchange scattering,and their magnetic and structural phasetransitions will be investigated. Second,non-magnetic materials: lattice distortionsin bulk material and real (i.e. non-UHV)

surfaces and interfaces.Examples from commissioning

experiments are shown below. Figure 2presents the intensity of the (0,0,2+q)magnetic satellite of holmium as afunction of energy at the LIII resonance.The resonant enhancement at the MIVedge of uranium is exemplified byFigure 3 for a mixed UAs/USe singlecrystal, just below its Neel temperature.The beamline appears to be workingaccording to specification.

Azimuthal angle (micro-radians)

-6 -4 -2 0 2 4 6

99.8 %

99.6 %

99.4 %

99.2 %

20

10

0

-10

10 %

5 %

0 %

-5 %

-10 %

A

C

PL

fP

B

Fig. 1: Curve A is thelinear polarisation PL

calculated frommeasurements of

Bragg peaks observedat a scattering angle

of 90° in thehorizontal plane at the

XMaS beamline as afunction of the

azimuthal angle.The degree of circular

polarisation, PC, isshown in curve B andfP (= PC/(1-PL) is thefactor that enhancesthe ratio of magneticto charge scattering.

The benefit of ahighly polarised

source is self evidentand the curve showsthat an angle of 2-3

microradians isoptimum for the white

beam experiment.

Fig. 2:Resonantenhancementat the LIIIabsorptionedge ofholmium.

Fig. 3: Theuranium MIV

resonance for aUAs/USe

sample, justbelow TN.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

49

T

T

IF (FRENCH BEAMLINE) ON BM32: INTERFACES

The word «interface» has a very wide meaning, and interfaces may be found in nearly everything,

from materials to living beings. Hence a large number of researchers from very different scientific

fields have been interested in the project from the beginning, which is illustrated by the large diversity

of experiments that have been performed concerning new materials, soft condensed matter,

environmental science or biology.

Because of its very versatile design, theGM instrument is a user-orientedfacility that fits the needs of a widecommunity. The research on the «GM»diffractometer has been balancedbetween studies of « soft » and « hard »condensed matter.

In the field of soft-condensedmatter, many experiments have beendevoted to the study of monolayers suchas amphiphilic films on liquids [1] andof free-standing liquid crystal films [2],intended to test theories on the elasticproperties of membranes with fullcalculation of the fluctuation spectrum,from macroscopic to molecular sizes,including various kinds of interactions.

Liquid crystal films supported by solidsubstrates like MoS2 have also beeninvestigated to determine the structuralevolution with temperature andcorrelate it to the observations of atomicforce microscopy, thus providinginsight on the mechanisms of inter-molecular and substrate/moleculeinteractions. These studies open uplarge perspectives on the influence ofdimensionality and substrate adhesionon order phase transitions in liquid-crystals. Other experiments have beendevoted to the 2D transitions of alcohollayers adsorbed at the air/waterinterface, with a particular attention tothe influence of chirality, opening the

he beamline has been focused on twocomplementary techniques: Glancing x-ray diffraction, which probes the longrange order, and x-ray absorptionspectroscopy (XAS) which probes theshort-range order. This resulted in threeexperimental stations, one equippedwith a multi-purpose goniometer (GM),one dedicated to x-ray absorptionspectroscopy (XAS) experiments, andone devoted to surface x-ray diffractionin ultra-high vacuum called SUV.The two first instruments werecommissioned in 1994, and the last onein 1996. The details of the beamline andexperimental stations can be found inthe ESRF Beamline Handbook.

he move of biocrystallography fromD2AM to FIP has started in March1998. However FIP will come intofull operation only by the end of 1998.What can be foreseen at present isthe following: in September 1998,the beamline will be ready with thesame equipment and similar beamcharacteristics as it was on D2AM, andwe will start to invite selected externalusers for data collection. But with respectto D2AM, two new developments areprepared: 1) using cryo-cooling on thefirst monochromator crystal and 2) usingpiezoelectric actuators for bending a longU-shaped crystal. The first improvementwill allow the extension of the range of

accessible x-ray photon energy up to25 keV, and the second one should makethe focusing of the beam faster and moreefficient by sagittal bending. Newdetectors will also be installed on thebeamline. Next September we will startinstalling the cryo-cooling of crystal 1 ofthe monochromator and probably animage plate detection system. Later in theyear, a new CCD detector will beinstalled replacing the XRII-CCD usedon the D2AM station since 1995. Asanother improvement with respect toD2AM, the beamline will have its owncold room as part of the small crystalpreparation laboratory located on thebeamline itself.

The gain in beam time for datacollection will contribute to improve thequality of the collected data in two aspects,first more time for data collection, of course,i.e. better statistics, but secondly more timefor careful data collection preparation. It willalso allow, hopefully, the development oftwo time-consuming new applications,namely 1) kinetic studies of biologicalreactions produced in the crystals, bycollecting monochromatic data after flash-cooling of the crystals in intermediate statesof reaction, and 2) the fine study of the x-raydiffraction in proteins with heavy atoms nearthe absorption edge, which requirescollecting data at a high number ofwavelengths on the absorption edge.

FIP (FRENCH BEAMLINE)ON BM30:INVESTIGATION OF PROTEIN STRUCTURES

FIP is a French CRG beamline under construction at the BM30 front-end. It will be dedicated

exclusively to biological macromolecular crystallography. FIP will replace the biocrystallography

station at the CRG beamline D2AM, where it used only half of the available beam time, sharing the

space with materials science experiments.

M. Roth

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

50

way for studies on cholesterol. Anotherimportant subject has been the study ofthe growth and wetting properties ofadsorbed films of rare gas, with the firstdirect observation of a pre-rougheningtransition at the origin of the surfacemelting, which is in turn at the origin ofbulk melting [3, 4].

In the field of hard-condensed matter,numerous and diverse experiments havebeen performed, concerning thestructure of thin layers such assemiconductors (ZnTe, CdTe, GaN) ormagnetic alloys (FePd), oxide layers(ZrO2) or metal/oxide interfaces(Ag/MgO). Surface x-ray diffractionexperiments were also performed whenUHV is not necessary, like on theoxidised surface of gold, thus bringingnew insight to the physics of metallicoxidation. Less classical experimentssuch as Refl-EXAFS on NbSe2 layersor resonant x-ray magnetic diffractionon ErFeB magnets were also performed.

The SUV station allows reflectivity,surface x-ray diffraction and absorptionto be performed in ultra-high vacuum(UHV), either on clean surfaces or onthin films. Most experiments have

consisted in studying the evolutionduring their growth or during anneals ofthe structure and morphology of thinmetallic films elaborated in situ ondifferent kinds of substrates, metallic orrefractory, with a thickness varying froma fraction of a monolayer to 100 Å ormore. An example is the combination ofthe three techniques to characterise ultra-thin metastable Pt-Co alloys formed onthe Pt(111) surface by different routes,either by co-deposition of Co and Ptor by annealing of a thin Co deposit,in relation to their properties ofperpendicular magnetic anisotropywhich make them promising candidatesfor high density magneto-optic recording[5]. The growth of the analogous system,Pt on Co(0001), was also investigated atdifferent temperatures [6], with the sameobjective of linking the structure andmorphology to magnetic properties.Another research program is devotedto oxide surfaces and metal/oxideinterfaces, which are of interest becausethey are involved in many applications,and because the theories to describethem are still under development. Modelsystems such as the surface of alumina[4] or of magnesium oxide and itsinterface with different metals like Ag,Pd or Ni [7, 8] have been investigated,with emphasis on the epitaxialrelationships, the growth mode and thelattice parameter relaxation during thevery early stages of deposition. Theintroduction of coherency defects suchas stacking faults or interfacial misfitdislocations and their reordering forlarge deposits and upon annealing werecharacterised in detail. The ultimate goalis to better understand, and improve, thebonding between such dissimilarmaterials. Other oxide surfaces such asNiO(111) [5, 9], are also studied becauseof their interesting magnetic properties.

The XAS station is now used by avery wide community, in biology,geochemistry, environment, electro-chemistry, catalysis, magnetism andmaterial sciences. All the groups towhich beam time has been attributedhave been able to realise theirexperiments (more than 25 publicationwith reviewers). Among them, let us citethree examples that are connectedto instrumental developments. Thedynamic focusing, allowing to keep a300 x 300 mm2 spot size on the sampleall over the EXAFS spectrum, has beensuccessfully used in many experimentssuch as in the high-pressure and high-temperature investigation on liquid

selenium near the critical point (390bars, 1620 °C), where a semiconductorto metal transition occurs [10]. Recently,a total electron yield detector with Hegas flow at atmospheric pressure hasbeen developed and successfully tested[11]. It was successfully used to analysethe evolution during annealing of thenanostructure of CoxAg1-x, NixAg1-xheterogeneous alloys exhibiting giantmagnetoresistance [12], as well as inan investigation of epitaxial (111) CoPt3films aimed at understanding themicroscopic origin of their perpendicularanisotropy [13].

A new germanium solid state multi-detector has recently been commissioned,which opens the way for studies of highlydiluted samples down to about 100 ppm.The next paper gives an example of its usein environmental science [14].

G. RenaudREFERENCES

[1] C. Gourier, J. Daillant, A. Braslau, D. Luzet,C. Blot, G. Grübel, J.F. Legrand, G. Vignaud,M. Alba, F. Rieutord and A. Gabriel, ESRFNewsletter 25, 34-37 (1996).[2] W.H. de Jeu, E.A.L. Mol, G.C.L. Wong, J.M.Petit and F. Rieutord, ESRF Newsletter 29, 10-12(1997).[3] F. Rieutord, R. Simon, R. Conradt andP. Müller-Buschbaum, Europhys. Lett. 37, 565(1997).[4] see ESRF highlights 1995/1996, p33 and 35.[5] An article on this subject will be found in theforthcoming publication of ESRF highlights.[6] H. Bulou, A. Barbier, G. Renaud, B. Carrière,R. Baudoing-Savois and J.P. Deville, Surf. Sci.377, 90 (1997).[7] O. Robach, G. Renaud, A. Barbier andP. Guénard, to appear in Surf. Rev. Lett. (1997)and O. Robach, Phd thesis, University ofGrenoble, 1997.[8] P. Guénard, G. Renaud and B. Villette,Physica B 221, 205 (1996) and P. Guénard, PhDthesis, University of Grenoble, 1996. [9] A. Barbier and G. Renaud, to appear in Surf.Sci., 392 L15-L20 (1997).[10] Y. Soldo, J.L. Hazemann, D. Aberdam,E. Pernot, M. Inui, J.F. Jal, K. Tamura, J. Dupuy-Philon and D. Raoux, J Phys. IV, C2-983 (1997),and accepted in Phys. Rev. B.[11] C. Revenant Brizard, J.R. Regnard,J. Mimault, D. Duclos and J.J. Faix, J. Phys. IV,C2-325 (1997).[12] C. Revenant Brizard, J.R. Regnard,J. Mimault, O. Proux, B. Dieny and B. Mevel,J. Phys. IV, C2-1111, 1997.[13] C. Meneghini, M. Maret, M.C. Cadeville andJ.L. Hazemann, J. Phys. IV, C2 1115-1117.[14] A. Manceau, M. C. Boisset, G. Sarret,J.L. Hazemann, M. Mench, Ph. Cambier, R. Prost,Env. Sci. & Technol., 30, 1540 (1996) .

The project of a French CRG beamlinededicated to the study of interfaceswas initiated in the framework of theFrench «Programme d'accompagne-ment de l'ESRF». It was supported byseveral laboratories from the NationalCenter For Scientific Research(CNRS, Laboratoire de Cristallo-graphie, Grenoble), the AtomicEnergy Commission (CEA, DRFMC,Grenoble and DRECAM, Saclay)and the University of Grenoble(Spectrométrie Physique). Part of thefinancial support also came from theIsère department and the Rhône-Alpes region.It was the very first CRG project to bepresented - and accepted - by theESRF, in early 1990. The IF beamline,one of the first available to users,came into operation in 1994.Experiments have been performedsince, with 1/3 of the available beamtime being allocated by Europeancommittees through the usualapplication for beam time at theESRF, and 2/3 being allocated by aFrench committee through a similarprocedure working at the French level.

SOME HISTORY

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

51

ENVIRONMENTALSCIENCE

G. SARRET1, A. MANCEAU1 AND

J.L. HAZEMANN1,2

1 LABORATOIRE DE GÉOPHYSIQUE INTERNE ET

TECTONOPHYSIQUE, GRENOBLE, FRANCE

2 LABORATOIRE DE CRISTALLOGRAPHIE CNRS, GRENOBLE, FRANCE

In recent years, environmentalproblems such as pollution of air and soilhave become increasingly important,even for everyday life. For these reasons,the ESRF has been involved in studiesconcerning environmental science. Inparticular, absorption experiments are anindispensable complement to laboratorytechniques and, sometimes, the onlymethod able to bring unambiguousanswers to environmental problems.

Recently, new possibilities to carryout absorption experiments on highlydiluted samples have been available onbeamline BM32. It is possible to study adiluted sample down to about 100 ppm,which is a very important target,allowing the study of samples whichwere, to date, inaccessible. The detectoris quite user-friendly: it is possible in aneasy way and quite rapidly to change theenergy selected intervals of detection ofthe fluorescent photons.

Furthermore, the dynamic focusing ofthe x-ray beam allows a 300-300 µm2 spotsize to be kept on the sample all over theEXAFS spectrum. This permits us, firstly,to considerably enhance the intensity ofincident photons on the sample, which isessential in the case of highly dilutedsystems, and, secondly, to study smallsamples even in grazing incidence.

DIRECT DETERMINATIONOF THE SPECIATION OFHEAVY METALS IN SOILS

BY EXAFSThe input of heavy metals in the

environment, and specifically in soilsand industrial or domestic urban wastes,endangers living organisms. Assessingthe risk associated with their presence isa prerequisite for designing recoverytechniques of contaminated sites andpreventing future contamination in thecase of landfills. In either case, the riskassociated with the presence of heavymetals depends primarily on theirspeciation. The notion of speciation is

taken in its broadest sense and includesmetal characteristics, such as electronicstructure (oxidation state, nature ofchemical bonds) and chemical state: itsassociation with the organic or inorganicfraction, the nature of the functionalgroups to which the metal is bonded,fixation by minerals (clays, oxides...)and the crystal chemical mechanism ofthis fixation (adsorption vs. latticesubstitution), or, more simply,precipitation of the pollutant as a salt(sulfate, carbonate, phosphate...), anoxide or a silicate. This chemical statedetermines the intrinsic toxicity of thehost matrix (soil or waste), as well as themobility of metals in the environment toeither surface and ground waters, livingorganisms (plants, micro-organisms,meso-fauna) or the atmosphere.

Determining the speciation of heavymetals is a difficult task because of theirrelatively high dilution and the structuraland chemical complexity of hostmaterials. Classical approaches are basedon the «selective» or the «sequential»dissolution of soils components and canbe classified as possibilities offered byEXAFS spectroscopy for determiningthe speciation of metals. In contrast toprevious methods, EXAFS is a «directmethod» since it can be applied to non-disturbed, polymetallic and polyphasicmaterials. The potential of the EXAFSspectroscopy method will be illustratedwith the determination of the speciationof heavy metals in lichens contaminated

by different sources of pollution of leadand zinc.

UPTAKE MECHANISMSFOR THE FIXATION OFHEAVY ELEMENTS BY

LICHENSLichens are currently used as

biomonitors of atmospheric and soilpollution. The absorbing capacity oflichens for different heavy metals hasbeen widely studied, but little is knownabout the localisation and retentionmechanisms of metals in the thallus.Three different uptake mechanismsare supposed to exist: (1) entrapmentof metal-rich mineral particulatesoriginating from atmospheric aerosols(mineral phase) (2) extracellularcomplexation with polysaccharides ofthe cell walls, oxalic acid or lichenicacids (3) intracellular uptake bycomplexation with cysteine-containingproteins (metallothioneins) (Figure 1).

EXAFS measurements oncontaminated lichens are a difficulttask because (i) metals are generallydiluted (500-1000 mg/kg), and (ii)lichens may absorb a variety of metals,thus creating a multi-element system.This study has been possible thanks tothe Canberra multidetector and thedynamic focusing allowing a detectionlimit for trace elements presently

Fig. 1: Threedifferent uptake

mechanisms:(1) entrapment of

metal-rich mineralparticulates

(2) extracellularcomplexation,

(3) intracellular.

COLLABORATING RESEARCH GROUPS

ESRF - APRIL 1998

52

comprised between 100 and 500 mg/kg.Two lichens from two French siteswere exposed to different pollutions:Xanthoria parietina, a saxicolous lichengrowing on rocks near a tetraethyland tetramethyl lead factory, andDiploschistes muscorum, a terricolouslichen collected on Pb-Zn tailings near aZn extraction plant.

Comparisons of EXAFS spectra forlichens and for reference compounds(Figure 2) show that in the two lichensstudied, metals are not included inmineral phases, but are complexed byorganic ligands. The comparison withPb or Zn-cysteine reference compoundsenable us to reject the intracellulartrapping mechanism for the two lichens.Therefore, three possible complexingagents could be involved in metalaccumulation within the lichens:lichenic acids, macromolecules of thecell wall, or oxalic acid exdudates.• Diploschistes: in this lichen, Pband Zn (Figure 3) were found tobe bioaccumulated by the samemechanism, i. e. by precipitation of Pb,

Zn oxalate. The production of oxalicacid is probably enhanced by thepollution stress.• Cladonia: in this lichen, Zn iscomplexed by lichenic acids but notwithin the cell wall.

This study reveals that in bothlichens, cells are protected from toxicityby extracellular complexation of heavymetals, but the strategies differ:in Diploschistes, Pb and Zn areaccumulated through an enhancedexcretion of oxalate, which precipitatestoxic elements such as insoluble salts,whereas in Xanthoria, Pb is complexedto carboxylic groups of the fungalcell walls. We conclude that hyper-accumulation of metals results from areactive mechanism of organic acidproduction, whereas metallo-toleranceis achieved by a passive complexationto existing functional groups.

REFERENCES

A. Manceau, J.C. Hargé, G. Sarret,J.L. Hazemann, M.C. Boisset, M. Mench, Ph.Cambier, R. Prost, 1996, Direct determination

of heavy metals speciation in soils byEXAFS spectroscopy, Proceedings of thethird International Conference on thebiogeochemistry of Trace Elements, 15-19May, Paris, in press.A. Manceau, M.C. Boisset, G. Sarret,J.L. Hazemann, M. Mench, Ph. Cambier,R. Prost, 1996, Direct Determination of leadspeciation in soils by EXAFS spectroscopy,Env. Sci. & Technol., 30, 1540, 1552.G. Sarret, A. Manceau, J.L. Hazemann,A. Gomez, and M. Mench, EXAFS study of thenature of zinc Complexation sites in humicsubstances as a function of Zn concentration(1997), Journal de Physique IV, C2 799-802.G. Sarret, A. Manceau, D. Cuny, C. VanHaluwyn, S. Déruelle, J.L. Hazemann, Y. Soldo, L. Eybert-Bérard, J.J. Menthonnex:Mechanism of lichen resistance to metallicpollution (1997) submitted to EnvironmentalScience and Technology.G. Sarret, A. Manceau, L. Spadini, J.C. Roux,J.L. Hazemann, Y. Soldo, L. Eybert-Bérard,J.J. Menthonnex: Structural determination ofZn and Pb binding sites in Penicilliumchrysogenum cell wall by EXAFS spectroscopy(1997) submitted to Environmental Scienceand Technology.

Fig. 3:Comparison

of Pb LIIIedge EXAFS

spectra fora lichen

and somereference

compounds.

Fig. 2:Comparisonof Zn K edgeEXAFSspectra forthe twolichens andsomereferencecompounds.

RESEARCH & DEVELOPMENT

ESRF - JULY 1997

53

Real-time x-ray imaging experiments,with typical evolution times in the0.1 second range, are now possible atseveral ESRF beamlines. The FRELON(Fast REad-Out LOw-Noise) CCDcamera, with 14-bit dynamic range and1024 x 1024 pixel area, developed at theESRF [1] is used on many of thesebeamlines. The standard solution toread this type of CCD camera data isusing a SDV card directly plugged inthe Sun workstation, and unified SPECsoftware. But this solution has not beenable until now to take advantage of thefull read-out speed of the FRELONcamera (20 MPixels/s).

We present, in this short report, real-time phase images recorded on the ID19('Topography') beamline, using a VMEsystem based on the VCCD3 card whichallows one to record about 20 images /second on the complete area and with thefull dynamic range. This system wasdeveloped in the Computing Services [2].

THE VCCD3 CARDThe VCCD3 VME card has on-board

memory for image rebuilding, organizedin two banks. While one bank iswritten/filled with the incoming pixelsfrom the FRELON camera (new image),the other bank (previous image) isread/flushed into the VME memory card.The VCCD3 card allows:• reading out the FRELON camera dataover all 4 optical fibers to achieve theshortest possible read-out time(20 images/seconde).• displaying a live image on a videodisplay for monitoring purposes.• summing up the data from theconsecutive images of shorter durationwhen required.• acquiring data in a continuous modeusing the VME memory as a ring buffer;a pre-defined number of images can bestored after an interesting event (trigger).The VCCD3 card stores the acquiredimages in a memory card, while BIT3adaptor cards and fast link (25 MBytes/s)allow the online display program to berun on the Sun workstation and the datato be transferred in an efficient way. Onthe software side, the dedicated OS-9device driver and server, as well as a

special SPEC application, were used toboth control the acquisition and totransfer the data.

FIRST RESULTSThe FRELON camera was equipped

with an optics leading to a pixel size of 10µm, a compromise allowing us to havereasonably short exposure times whileretaining a fair spatial resolution. Anenergy of 20 keV was selected by thedouble-crystal (Si 111 in Bragg sym-metrical geometry) vertical monochro-mator. For the first tests with a pureabsorption object (stainless steel tip), thesample was located just before the x-rayconverter (a Gd2O2S:Tb scintillatorscreen). For the final series of images ofphase objects (gas bubbles in beer), theywere placed at around 2.5 m from thecamera in order to obtain the phase imagethrough propagation (see papers byBuffiere et al., and Baruchel et al., in thisNewsletter, pages 18 and 20). At thisdistance, we were in an edge detectionregime, where black/white contrastappears at the border of the gas bubbles.

The images were corrected bysubtracting the 'dark current' (withoutbeam), and the 'flat field' images (withbeam but without sample). A «fast»shutter (7 mm x 7 mm) was used to cut offthe x-ray beam during the read-out time.

We tried to recreate the conditionsof a real experiment in which we wouldhave phase objects (thus, invisible inabsorption) which move in someunpredictable way and whose behavior

we wanted to track at a given moment.We found all these characteristics in thegas bubbles which appear inside a glassof beer (of course Champagne wouldalso have been great, but it is moreexpensive). We took a series of data onseveral samples, using an exposure timeof 50 ms which, together with the read-out time, yields a lapse of time of 100ms between two consecutive images.

Figure 1 shows, as an example, aseries of images recorded over a bubblesource inside the beer. The emergingbubble is recognized when it has adiameter of ~ 30 µm. We can perfectlytrack its movement as it grows up toø ~ 150 µm and its speed increases.

CONCLUSIONSBy using the FRELON camera at its

full read-out speed, as carried out with theVCCD3 card, we could for the first timeperform real-time phase-contrast imaging.The real-time display proved to be crucialfor the success of the experiment.

The implementation of a fast read-outsystem well integrated within the standardESRF hardware and software appears tobe essential for the real-time imageacquisition, which is of increasing interestfor many users.

REFERENCES

[1] J.C. Labiche, J. Segura-Puchades, D. vanBrussel and J.P. Moy, FRELON camera: FastReadout Low Noise, ESRF Newsletter N° 25,p 41-43, March 1996.[2] C. Hervé, J. Cerrai, VCCD3 specification,version 0.4, CS/EL/95/0x.

ESRF - APRIL 1998

REAL-TIME PHASE-CONTRAST IMAGINGP. CLOETENS1, J.I. ESPESO1 AND F. SEVER2

1 ESRF, EXPERIMENTS DIVISION

2 ESRF, COMPUTING SERVICES DIVISION

t = 0 s 0.1 s 0.2 s 0.3 s 0.4 s 0.5 s 0.6 s 0.7 s

Exposure time = 5 x 10–2 s1 mm

4 m

m

Fig. 1: Phase-contrast images showing the formation and evolution of gasbubbles inside a glass of «white» beer (a Belgian beer which is opaque).

RESEARCH & DEVELOPMENT

ESRF - JULY 1997

54

Magnet alignment errors areresponsible for different types ofperturbations in the ESRF storage ring. The main effects are:• Quadrupole magnet positioning errorsinducing closed orbit distortions,• Quadrupole tilt error inducing emittancecoupling into the vertical plane.These perturbations are normallycompensated by magnetic correctors(horizontal or vertical steerers, skewquadrupole corrector magnets). Thanksto the good initial alignment of themachine, these correctors are used at asmall fraction of their nominal strength.Alternatively the remote controlledmotorised jacks supporting the magnetgirders can also be used to minimisesome of the misalignment errors:• Vertical quadrupole positioning• Quadrupole tilt (though this motion iscorrelated with an undesired horizontaldisplacement of the magnets).The interest of correcting mechanicallythese errors may be:• The analysis of the correction givesindications on the residual errors(amplitude, distribution) after thealignment of the machine,• Ensuring a certain level of correctioncould permit the reduction of the strength

of the magnetic correctors and at thesame time improve their resolution. Thiswould be beneficial for the fineadjustment of the beam position.Several experiments have beenperformed recently to check these twocorrection procedures.

The remote control of the jacks,together with the Hydrostatic LevellingSystem (HLS), which cross-checks theeffective girder displacement, enablethese tests to be performed with thetunnel closed and with 5 mA of storedbeam. The net advantage is that the effect(beneficial or not) of the girderdisplacements on the electron beam isassessed on-line. This beam-basedalignment technique is a long awaiteddream of accelerator physicists aroundthe world. The different experimentswere performed within a few hours andthe girder’s position and beam trajectorywere put back at the end of the tests.

CONTROL OF GIRDERMOVEMENTS

Figure 1 shows the position of the HLS andjacks on the G10, G20 and G30 girders.Longitudinal tilt motion is a rotation about

the middle jack on a girder inthe sense of the travel of thebeam. Radial tilt motion isabout the center of the girder inthe sense perpendicular tothe beam travel. Verticalmovements are made on theG10, G20 and G30 girders.Jack movements are calculatedfrom longitudinal and radial tiltvalues issued from calculationsof beam parameters. Thesetilt values are translatedinto movements for the threejacks under each girder.Corresponding expected HLSreadings at the jack positionsare also calculated and used as acontrol for these movements.The difference between the jack

movement and the expected HLS readingis the precision of the movement. Inpractice, the HLS readings must beprocessed to eliminate the wave effect dueto the motion of water in the pipes, and thepossible discontinuities in the readingscreated by blockages in the water system.For a 40 µm peak movement of the jacks,the residual standard deviation is 1.3 µmwhich represents both the precision of the288 jack movements and the naturalevolution in time of the storage ringgirders.

VERTICAL BEAM CLOSEDORBIT DISTORTION

Calibration of girder displacementsEighteen jacks were moved

independently by 10 µm. The motion waschecked by deducing the displacementfrom the beam position readings on theelectron Beam Position Monitors (BPM)all around the storage ring. There is anagreement of better than 1 µm between therequested jack movement, the HLSreadings and the beam response. Closed orbit correction using girderdisplacements

The response matrix of the verticalclosed orbit to girder motion wascomputed for the theoretical machine. Insimulation analysis, it was confirmed thatpure longitudinal rotation of girders wasmore efficient than pure translation. It wastherefore decided to correct the machineusing only girder rotations. The correctionmethod was exactly the same as the oneused for magnetic steerers (SVD method).The procedure used in the test was:• Start from a perfectly corrected

machine,• Reduce the number of steerer correction

vectors so that the vertical orbit blowsup significantly (zrms < 300 µm),

• Measure the vertical closed orbit andcompute the mechanical correction,

• Apply the mechanical correction,• iterate.The results were obtained after four

ESRF - APRIL 1998

MECHANICALLY INDUCED INFLUENCES

ON THE STORAGE RING BEAM

D. MARTIN1 AND L. FARVACQUE2

1 ESRF, TECHNICAL SERVICES DIVISION

2 ESRF, MACHINE DIVISION

Jack

Jack

HLS

HLS

Radial Sense LongitudinalSense

Fig. 1: position of the HLS and jacks on atypical storage ring girder.

RESEARCH & DEVELOPMENT

ESRF - JULY 1997

55

iterations (see Table 1).Figure 2 shows the displacementscorresponding to this partial alignment:The full cancellation of the steerer currentscould not be tested because of lack of time,but should be straightforward. The ratherlarge residual orbit is due to the choice ofa limited number of correction vectors toavoid too large movements of girders.

HORIZONTAL/VERTICALEMITTANCE COUPLING

The horizontal/vertical betatron couplingis responsible for the major part of thevertical emittance of the storage ring. It isdefined as the ratio k of the vertical to thehorizontal beam emittances:

εz = k εx

In a perfect machine coupling doesnot exist. It is a result of magnetimperfections and alignment tilterrors. Since the brilliance is inverselyproportional to the coupling, itsreduction is a way of optimising theperformance. This is usually done bypowering skew quadrupole correctors.Alternatively, coupling can be variedby tilting girders in the radial sense.The horizontal displacement linkedto the tilts also induced undesiredhorizontal closed orbit distortions whichwere corrected by the standard orbitcorrection system.Calibration of a harmonic tilt

The coupling is mainly sensitive tothe excitation of the resonance close tothe betatron tune difference (νx - νz = 25in the case of the 4 nm lattice). So asystematic transverse tilt was applied tothe girders on a corrected machine,according to this 25th harmonic:The value of the peak angle was variedbetween 0 and 1 mrad and the couplingwas measured using the pinhole camera.The results show a large deviation(factor 2 to 3) from the prediction of thetheoretical model. This discrepancy isnot understood at this point.Correction of the main harmonic withthe storage ring girders

This was applied on the low βzoptics for which the tune difference is

νx-νz = 22. A harmonic 22 tilt wasapplied to a machine without correctionsand was varied experimentally to cancelthe excitation of the coupling resonance:

rms tilt angle Phase Coupling(mrad rms) (°)

0 0 34 %0.21 19 4.5 %

The residual coupling value issimilar to what can be obtained withmagnetic correction of a singleresonance. However, the rms tilt value of0.21 mrad (see Figure 3) introduced inthe machine is very large. It cannotrepresent the compensation of a residualharmonic component with such anamplitude in the girder alignment. Itprobably compensates another couplingsource (individual positioning of magnetson girders, magnetic tilt angle,…).

CONCLUSIONSeveral experiments have been

carried out using high precision jacksinstalled under the SR girders relating

mechanical motion to vertical closedorbit and coupling: • the calibration of the translation of onegirder, • the complete correction of the machineby movements made to imitate theaction of the steerers,• the calibration of a harmonic excitationof a coupling resonance as a function ofgirder tilts,• the compensation of the main couplingresonance.

The correction of coupling shows asensitivity of the machine much higherthan predicted and movements largerthat what could be derived fromalignment tolerances.

The correction applied for the verticalclosed orbit agrees perfectly with themodelling and results in motionscompatible with the residual alignmenterrors. This may be applied for the nextmachine realignment (performed withbeam after the next winter shutdown) andcould allow a further reduction of the rmsvertical closed orbit errors. It could alsoallow the achievement of a low rms orbit(~ 200 µm) without any magneticcorrection which would be a record forsuch a high focusing storage ring.

ESRF - APRIL 1998

Til

t (m

rad

)

0.3

0.2

0.1

0

-0.1

-0.2

-0.3

Fig. 3: Tilt movements used for coupling correction.

Nb. vectors rms. orbit before rms. orbit after rms. steererleft on steerers movements movements strength

(µm) (µm) (mA)

1st iteration 32 269 191 76last iteration 3 493 215 21

dH (

µm)

200

150

100

50

0

-50

-100

-150

-200

Fig. 2: Girder displacement after correction.

Table 1

machine circumference

machine circumference

The ESRF Newsletter is published by the European Synchrotron Radiation FacilityBP 220, F38043 Grenoble cedex

Editor: Dominique CORNUÉJOLS Tél (33) 4 76 88 20 25Dépôt légal : 1er trimestre 1998. Printed in France by Repro-Express. Layout by Pixel Project.

On ID14, a beamline for proteincrystallography.

In the optics lab.

Theatrical itinerary«Lumières»from 20 to

31 March 1998

6 actors and 1 musiciancreated a new kind of visit at the

ESRF, mixing Science and Art.(See page 11)

EventsEvents

Open Days14-15 March 1998

About 2000 people visitedthe ESRF on this occasionand had the opportunity tomeet with the ESRF staff.(See page 11)