annual report 2003 zwe frm-ii · 2015-03-04 · frm-ii takes part and will welcome its ﬁrst eu...

ANNUAL REPORT 2003ZWE FRM-II

iv Contents

Contents

Director’s Report 1The Nuclear Start-up of the FRM-II. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

The Year in Pictures 3

1 Central Services 61.1 Detector and Electronics Laboratory. . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.2 HELIUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.3 IT-Network Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.4 Neutron Guides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .131.5 Supermirrors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151.6 Sample Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161.7 Software Group. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171.8 NICOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201.9 Radiation Protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

2 Diffractometers 242.1 Biodiffractometer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242.2 HEIDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262.3 MatSci-R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282.4 REFSANS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .302.5 REFSANS Detector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .322.6 RESI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .342.7 SPODI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .352.8 STRESS-SPEC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36

3 Spectrometers 383.1 MIRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .383.2 NRSE-TAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393.3 PANDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423.4 PUMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .433.5 RESEDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .463.6 RSSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .473.7 TOFTOF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50

4 Fundamental Physics 524.1 MAFF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524.2 MEPHISTO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .564.3 Nepomuc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58

Contents v

5 Irradiation and Radiography Facilities 615.1 ANTARES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .615.2 NECTAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .635.3 Irradiation Facilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65

6 New Developments 686.1 Novel Neutron Image Plates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686.2 Neutron Dosimetry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .716.3 Image Deconvolution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .726.4 Coded Apertures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .736.5 Phase Contrast Radiography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746.6 NRSE coils for MIRA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .756.7 Dynamic Neutron Radiography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776.8 Fuel Element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

7 Scientific Highlights 807.1 Pressure Induced Transitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807.2 Critical Exponents in MnSi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837.3 Spin waves in EuS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .857.4 Submicrometer X-Ray Tomography. . . . . . . . . . . . . . . . . . . . . . . . . . . . 877.5 Stroboscopic Neutron Imaging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

8 Events, People and Figures 918.1 Workshop on Neutron Scattering. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918.2 Public Relations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .928.3 Committees. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .948.4 Staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .978.5 Publications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101

Imprint 105

vi Contents

Director’s Report 1

Director’s Report

The Nuclear Start-up of the FRM-II

After a very long and diffi-cult procedure the FRM-II ob-tained its license for the nuclearstart-up and operation (3rd li-cense) in May 2003 from theNuclear Licensing Authority, theBavarian State Ministry of StateDevelopment and Environment(StMLU), now renamed to Bavar-ian State Ministry of Environ-ment, Health and Consumer Pro-tection (StMUGV). The firstdraft for this license was submit-ted already in August 2000 andhad to be agreed upon by the Fed-eral Ministry of Environment andReactor Safety (BMU). The finallist of questions from this federalministry was answered in Jan-uary 2003 by Siemens/TUM andfinally opened the way for the li-cense. The 3rd license includesinstructions for running the facil-ity and the condition to develop afuel element with reduced enrich-ment in235U.

The event of issuing the nu-clear license for FRM-II has beencelebrated by a visit of the Bavar-ian Prime Minister, Dr. Stoiber,the Bavarian Minister of Scienceand Art, Mr. Zehetmair andthe Bavarian Minister of Environ-ment, Dr. Schnappauf on June 4th

at FRM-II.With the 3rd license we imme-

diately initiated the delivery ofthe heavy water for the moder-

ator, the first fuel elements andthe converter plates. We receivedthe heavy water in June 2003 andthe fuel elements in July 2003as well as the primary neutronsource for the nuclear start-up.The heavy water was filled safelyinto the storage tank. The fuel el-ements were inspected and mea-sured by EURATOM and IAEAon its 235U content one day afterdelivery.

Prior to the nuclear start-up ageneral inspection of all reactorcomponents was required, sincethe starting of the systems in2001 was too far away due to thelong licensing procedure. Thegeneral contractor Siemens hadto reestablish its working teamfor the nuclear start-up. Thesystems were also checked inrespect to their long-term oper-ational behavior and were im-proved accordingly.

The nuclear start-up from firstcriticality to full power will startin early 2004 and end with acontinuous run at 20 MW inJune 2004. Important steps arethe calibration of the shut-downand control rods, the shut-downsafety proofs, the power den-sity distribution measurements,the neutron flux measurements,the power calibration and controland tests of the cooling systems.During this program the various

special facilities such as ColdSource, Hot Source, Converter,irradiation facilities and silicondoping device will also be takeninto operation. First neutronswill also be available to take intooperation most of the 19 beam po-sition instruments, currently un-der construction. It is expectedthat 14 of these instruments in-cluding the positron source willreceive their first users with thestart of the second fuel cycle. Thenuclear start-up program endswith the burn-up of the first fuelelement. Based on a report on theexperience with the first fuel ele-ment the licensing authority is inthe position to permit the routineoperation of the FRM-II whichwill start with the 2nd fuel cycle.

Within the 6th frame work pro-gramme of the EU aNeutronMyon Integrated InfrastructureInitiative NMI 3 has been grantedto the most powerful neutron andmyon centers in Europe. FRM-II takes part and will welcome itsfirst EU supported users in 2004.

The German Federal Min-istry for Science and Technol-ogy starts with April, 1st 2004 anew 3 years period of grants de-voted to facilitate the access ofuniversity groups to neutron andsynchrotron facilities. Univer-sity groups involved in the instru-

2 Director’s Report

mentation of FRM-II will largelyprofit from this programme.

Construction work on theFRM-II site never ends. Withhelp of the Bavarian investmentprogramme in science and tech-nology, HighTech Offensive Bay-ern, a center for industrial appli-cations of the FRM-II neutrons iscurrently under construction. InAutumn 2004 the first industrialusers will settle in the new labsand offices.

The use of polarized neutronswill be an essential tool for neu-tron research in the future. Asa central service for the instru-ments a facility to produce po-larized3He gas as filter mediumfor polarized neutron beams hasbeen ordered in May 2003. A keyready system will be delivered inApril 2004 at the site of the FRM-II with a capacity of 30bar·litersper day.

With the issue of the nu-clear license our advisory boardshave also changed. On Octo-ber 9th, the "Projektbegeleiten-der Beirat", our advisory boardduring the constructions phasemet for the last time. Prof.Dr. Tasso Springer who has ledthis board since 1994 has beenhonored by the Karl-Max-von-Bauernfeind medal. At the sametime the new advisory board, nowcalled Strategierat did have itsconstitutive meeting under thepresidency of Prof. Dr. GernotHeger.

Erich Steichele presenting neutronguides at the FRM-II

At the end of October 2003 Dr.Erich Steichele retired after 37years of work devoted to neutronresearch at the Technische Uni-versität München. Major projectsof Eric Steichele have been a neu-tron time-of-flight diffractome-ter for high resolution studiesand the development of neutronguide systems and neutron opti-cal devices. At the FRM-II heworked from the very beginningon the design of the secondarysources and beam tubes as wellas on the neutron guide system.For the new instrumentation heworked on the radiography sta-tions and the reflectometer where

he designed and realised the firsttwisted neutron guide. We wishhim all the best for a long andprosperous retirement.

We commemorate the death ofJürgen Didier (Framatom/ANP)on Febuary, 2nd 2004. He wasinvolved already in the conceptand planing of the FRM-II andtook over the responsibilities asproject leader in 1994 until 2002.We appreciated him as a verycompetent and always friendlyminded partner. Even in compli-cated negotiations he was willingto find a good compromise lead-ing to a reliable solution withoutloosing his fundamental humor.We will keep pleasant memoriesof Jürgen Didier with building ofour reactor.

Jürgen Didier

Guido Engelke Winfried Petry Klaus Schreckenbach


The Year in Pictures

16.1.03The group of french reactor operateurs (OEA)visits the FRM-II

10.3.03Parliamentary group from CSU with Dr. MMayer(MP) and J. Singhammer(MP)

4.4.03A. Berg(MP) visits the FRM-II

7.4.03J. Singhammer(MP) visits the FRM-II

4 The Year in Pictures

20-22.5.03Jahrestagung Kerntechnik in Berlin

4.6.03Bavarian Prime minister Dr. E. Stoiber arrives atthe FRM-II

4.6.03Bavarian Prime minister Dr. E. Stoiber and Prof.W. Petry

4.6.03 W. Schnappauf, bavarian minister for environ-ment, during his talk at the FRM-II

4.6.03H. Zehetmair, minister of science, talks at theFRM-II

4.6.03 Prof. Dr. W. A. Herrmann (President of theTUM)


1.10.03B. Bigot (Haut Commissaire à l’Energie Atom-ique), Reg.Dir. B. Schmid (StMWFK) and Min.Dir.Dr. H. Schunck (BMBF) visit the FRM-II

9.10.03Prof. Hermann awards Prof. T. Springer withthe Karl-Max-von-Bauernfeind medal for his long-standing support of FRM-II

30.10.03Production of the internal safety instructionfilm

30.10.03The FRM-II building in October 2003

6 1 Central Services

1 Central Services

1.1 Detector and Electronics Laboratory

K. Zeitelhack 1, A. Kastenmüller 1, S. Egerland1, D. Maier 1, M. Panradl 1

1 ZWE FRM-II, TU München

Corresponding to the progres-sive commissioning of the scien-tific instruments at FRM-II themain tasks in 2003 were to giveservice and assistance. In parallela few research and developmentprojects could be pursued and theupgrade of the laboratory infras-tructure continued.

Detectors

For most of the neutron scat-tering instruments we took careof inspection and adjustment ofdetectors and readout electron-ics purchased. An importantpart of this job was the qual-ity control of 605 squashed3He-detectors (filling gas: 3He/CF4

97/3; p=10 bar) installed at theTOF-TOF-spectrometer. De-veloped with Eurisys-Mesuresthese detectors show an im-proved charge collection due tothe CF4 admixture. This shouldresult in a better homogeneityand improved TOF-resolution forthermal neutrons. Fig. 1 showsthe pulse height spectrum of ther-mal neutrons from a252Cf-sourcerecorded with the new detec-tor type 120NH40RTF using fastreadout electronics (0,5µs shap-ing) in comparison to a conven-

tional Dextray type 90NH40Recsquashed series detector.

The detectors were deliveredin 5 batches within the periodof a year. Quality control in-cluded verification of mechan-ical dimensions using a semi-automatic measuring bench de-signed by our group and record-ing of pulse height spectra usingthermal neutrons from a252Cf-source for every single detector.

All detectors kept within themechanical specifications and re-vealed good homogeneity ingas gain and energy resolution

0 200 400 600 800 1000 1200 1400 1600 1800 20000

200

400

600

800

1000

90NH40rec

120NH40RTF gain: 35

20 %

15 %

10 % pulse shaping

tsh

= 0,5 m s

coun

ts

pulse height [channels]

Figure 1.1:Comparison of the pulse height spectrum of thermal neutronsrecorded with a 120NH40RTF and a conventional 90NH40Rec detector

independent of the batch. The rel-ative deviation (1σ ) in gain of thecomplete detector ensemble wasdetermined to 2.4 %. Three detec-tors showing a significant devia-tion could be proven to be leakyand were replaced.

Instrument Control andReadout Electronics

Installation and commission-ing of components for instru-ment control or detector read-out systems took a centerstage in 2003. Commissioning

1 Central Services 7

of the remote controlled bias andHV-supply system for theTOF-TOF detectors and completion ofslow control components for theIntense Positron SourceNEPO-MUC or the experimental site forvery cold neutronsMIRA com-prising a vacuum & dosimetryinterlock, remote bias and mag-net power supply, fieldbus inte-gration, pressure valve and mo-tion control were some major is-sues. Beside these projects the as-sistance to instrument scientistsin all technical questions relatedto electronics was a main task.

Several neutron scattering in-struments have been equippedwith our FRM-II standard read-out electronics developed for3He- and 235U-fission detectors.It consists of a amplifier/w-discriminator unit with variableshaping time, a remote con-trolled 3HE 19"-crate for bias

and HV supply and instrumentspecific PCI/cPCI data aquisitionboards [1]. Meanwhile, we de-veloped a whole family of de-tector data acquisition boardsfor counting, Time-of-Flight mea-surement or pulse height analy-sis in neutron scattering experi-ments. These boards consist ofa commercial PCI/cPCI or VMEcarrier board [2] and an appropri-ate M-module designed for theapplication required. The M-module types available are

• a counter/timer/ratemetermodule. It carries four32-bit counters with opto-isolated inputs that can beset as master/slave and op-erated in absolute, preset,timer or ratemeter mode.In addition, it provides2 programmable I/O chan-nels

• a multichannel scaler mod-ule (MCS). It provides 8input channels, 1µs - 32 sdwell time, 8 ms - 150 htime span (8k full mem-ory depth), 32 bit memorycapacity. It can be oper-ated in internal or exter-nal trigger mode. Start oftime span can be delayedto Trigger from 0 - 18 h

• a multichannel analyzermodule (MCA). It pro-vides 4 input channels anda 12 bit peak-sensing ADC.It can be operated in gatedor self triggered mode. Aprototype module has beensuccessfully tested

Control software and driversare available for FRM-II instru-ment software TACO or LAB-VIEW.

[1] Zeitelhack K., Kastenmüller A., Maier D., Panradl M.annual report.[2] MEN mikro-elektronik, Germany.

1.2 Project HELIUS – Polarized 3He for Neutron Instrumentation

S. Masalovich1, G. Borchert 1, W. Petry 1


Polarized nuclei of helium-3possess very high spin-dependentneutron absorption efficiencyover a wide range of neutron en-ergy. Neutron spin filters (NSF)based on a dense hyperpolar-ized 3He gas may compete inpolarization efficiency with suchcommon devices as magnetizedsingle crystals, supermirrors orsoller guides. Although theselatter are rather simple in op-

eration, their applications arestrongly limited by acceptableneutron energy and scatteringangle range. By contrast, thebroadband neutron spin filter canbe sized and shaped in such away that it will meet just aboutall practical needs. Until recently,the main disadvantage of thistechnique was a limited possibil-ity to produce highly polarized3He gas. However, the progress

achieved during last years in thisarea nearly eliminates this limita-tion. At present the3He setupsat the University of Mainz (Ger-many) and at the ILL (Grenoble,France) feature a production rate30 bar·litre/day for highly polar-ized 3He gas with polarizationof around 65%. Both setups uti-lize the direct optical pumpingof metastable3He atoms in3Heplasma at about 1 mbar. After


subsequent compressing, the po-larized gas at a pressure of a fewbars may be collected in a de-tachable cell of a given size andshape and then transported to aneutron instrument. Experimentsperformed at the ILL showed ahigh efficiency of NSFs over awide range of experimental con-ditions.

On July 2002 the expert’smeeting ”Realization of a3He fa-cility at the FRM-II” has beenheld at Garching. The main out-come of this meeting was a gen-eral agreement about the neces-sity of a3He polarization facilityat the FRM-II. A way to obtainsuch a facility was seen in a closecooperation with partner labora-tories at Mainz and Grenoble.

Fig. 1.2 shows the most gen-eral scheme of the production ofa polarized3He gas and deliver-ing this gas for neutron experi-ments.

This scheme can be dividedinto three levels:

• production of dense polar-ized3He gas (3He facility)

• delivery of this gas to neu-tron experiments withoutlosing a polarization (trans-portation)

• the use of a polarizedgas in neutron experimentswith proper environmentconditions (neutron experi-ments)

The most crucial part in thisscheme is certainly the produc-tion of polarized gas. At presentthis facility is under constructionin Mainz and expected deliverytime is April-May 2004.

High pressure. Detachable cells

with pol. 3He

Low pressure.Pol. 3He

Vacuumsystem

Homogeneousmagnetic field

RFdischarge

Laser Opticalsystem

Compressor

Transport

Neutron experiments

Absolute polarizationmeasurement

Absolute polarization measurement+ monitoring

Targetcells

Figure 1.2:General scheme of the production of polarized3He gas

During last year all necessarypreparations were made at theFRM-II to provide this facilitywith required infrastructure:

• the construction of a spe-cial ”nonmagnetic” labora-tory for the 3He setup isstarted at the local area,

• the laboratory for the cellspreparation is built andpartly equipped.

In the context of the agreementfor cooperation Dr. Masalovich(TU München) stayed for tenmonths at the Johannes Guten-berg University of Mainz wherehe learned the theory of the op-tical pumping and practice ofthe production of a polarizedgas. During his visit a newmethod for the precise measure-ment of a dense gas polarizationhas been proposed in a close co-operation with the University ofMainz. The first test measure-ments showed good results.

This work will be continued inthe next year.

During two-month stay atthe ILL (Grenoble, France) Dr.Masalovich was able to learn thepractice of polarized3He produc-tion there as well as NSFs appli-cation for neutron instrumenta-tion. Polarized gas transportationto a neutron experiment site andmagnetic shielding for polarizedgas at an experimental site wereof particular interest. This visitand knowledge transfer becamefeasible due to the goodwill ofthe ILL and financial supportfrom the European Community.In general, this visit is consideredto be a first step in a new closecollaboration between the Tech-nical University of Munich andthe Institut Laue-Langevin.

After construction and installa-tion of the3He facility and prepa-ration of necessary infrastructurethe first neutron spin filters is ex-pected to be available in summer2004.


1.3 IT-Network Services

Jörg Pulz1, Felix Zöbisch1, Jakob Mittermaier 1, Roman Müller 1, Joseph Ertl1, Sebastian Roth1,Sebastian Krohn1, Christoph Herbster 1, Jörn Beckmann1


Overview

Due to a significant growth of thenetwork, central servers had to bereplaced and changes in the ar-chitecture of the network wheredone. Major efforts and achieve-ments of this year were:

• Integration of the ReactorOperations Network andthe Scientific Network intoone FRM-II Network

• LDAP Server as methodfor unified access to userinformation

• New Mail-Servers• New FTP-Server• Terminal-Server Farm put

into operation• Cabling of instrument net-

works• Updated version of the

Taco-Box distribution• Setup and testing of a

Document- and Workflow-Management System

The FRM-II network is sep-arated into subnetworks. Eachinstrument, administration, reac-tor crew and the servers havetheir own subnet. Guests willbe able to connect their laptopsto a special network which willallow Internet access but limitsthe access to internal networksfor non-authenticated users. Theseparation of the individual net-works is done by means of Ac-cess Control Lists (ACLs) on thecentral router. The connection be-

tween the FRM-II network andthe Internet is provided by a two-stage firewall. Public accessi-ble services are placed in theDemilitarized Zone (DMZ) be-tween both firewall stages. Thearchitecture of the FRM-II net-work is shown in Fig. 1.3.

Figure 1.3:Architecture of the FRM-II network.

Unified Access to UserInformation

Different Operating Systems re-quire different protocols to storeand exchange User Information.With a common username dif-ferent passwords have to be set


for individual systems. Upto now this was done by theGeNUA User-Manager, a set ofPerl Scripts running on Unix sys-tems. For a new server the scriptshad to be changed.

In order to construct a ro-bust multi-platform User Man-agement System, LDAP was cho-sen as the central informationrepository. For redundancy oneMaster Server holds the infor-mation and distributes them onSlave Servers. The Slave Serversoffer read-only access and re-place the Master Server in caseof failure. To add new informa-tion or to change already storedinformation access to the MasterServer is necessary. LDAP of-fers password protection and isable to limit access to parts of thedirectory based on role and userinformation. We decided to useOpenLDAP [1].

Figure 1.4:Hierarchy of User Information Systems and Protocols

New Servers for FTP andEmail

For older Unix systems, wereNIS is still required, NIS mapsare created for the LDAP repos-itory.

Furthermore the LDAP sys-tem is used for authentica-tion purposes using the Samba[2]systems (access of Windowsmachines to Unix servers) andRadius [3](external connections).

As already mentioned in sec-tion 1 the old Mail Server wasnot able anymore to handle theuser load. The old Server wasbased on the Post Office Pro-tocol (POP) where users down-load their Email from the serverand store the Email on their lo-cal machine. This is impracti-cal for users working on morethan one machine and also a se-curity issue, as POP sends pass-words unencrypted over the net-work. Therefore IMAP was

chosen as the protocol for thenew Mail Server. IMAP storesthe Mail on the server and allowsusers to create subfolders whichcould also be shared with otherusers. As the Email is stored onthe server a central backup is pos-sible and the user can access hisor her Email from all computersinside the network.

Major concern with Emailnowadays are malicious contentand spam. To cope with thisthe new server was equipped withVirus and Spam Filtering func-tionalities. Mail processing isshared between two servers. Theserver for incoming messages is“mailhost”. On this machine Post-fix handles Email. The Add-onAmavis [4] connects Sophie [5]as virus scanner with Postfix. So-phie is a daemon process whichuses the Sophos library to checkincoming mails. After the viruscheck is passed, SpamAssasin[6] performs header analysis, textanalysis, consults blacklists andthe Razor signature database tofilter out unsolicited bulk Email.Messages which have passed thetests are handed over to Post-fix for further processing. If itwas sent from a FRM-II com-puter Postfix finally checks withthe LDAP Server if the user ex-ists and rewrites the “From” fieldin the Mail Header accordingly.If the Emails final destination isinside the FRM-II network it ishanded over to the IMAP Serverfor further processing.

On the IMAP server Send-mail first checks with the LDAPServer, if the recipient exists andforwards the Email to the CyrusIMAP Server. Users can collecttheir Email directly from Cyrus


IMAP, if they have an IMAPClient and reside inside the FRM-II network. Users from, outsidecan use the Webmail interfaceprovided by Squirrel Mail [7].Squirrel Mail offers full HTTPSaccess to IMAP servers and al-lows to manage Sieve Scripts onthe Mail Server. A functionto change the user password inthe LDAP repository is also in-cluded. Authentication of usersis via regular IMAP login. TheApache Web Server (with mod-ules) offers direct authenticationby means of the LDAP Server. InFig. 1.5 the blue arrows showthe way, an Email takes throughthe different programs. The redarrows show authentication re-quests of the different servers andthe green arrows show user infor-mation requested from the cen-tral LDAP repository.

On our mail servers the datatransfer is limited to 10 MB perEmail. To exchange larger fileswith external partners an FTPserver was built. FTP is not a se-cure protocol and transfers pass-

Figure 1.5:The Mail system at FRM-II. Blue arrows show the way an Emailtakes through the server hierarchy. The red arrows show Authenticationrequests and the green arrows show User Information provided.

words unencrypted. To preventusers from sending their pass-words unencrypted through theInternet, our FTP server only ac-cepts authenticated logins frominside the FRM-II network andanonymous logins only from theInternet. Every upload is viruschecked and reported to the ad-mins. Files in the incoming andoutgoing section are removed au-tomatically after 7 days.

Roll-out of the WindowsTerminal Server Farm

The configuration and idea be-hind a Windows Terminal Serverwas already explained in lastyears report. Just as a reminder,our server farm consist of threedual-Xeon 2 GHz servers with 4GB RAM with Citrix MetaFrameXPa. Standard Software is in-stalled on all three machines.User can either connect a fulldesktop session or open single ap-plications. If the user requestsjust an application a scheduler de-cides to which of the three

servers the user is connected.In case of full desktop sessionsthe user can decide to whichof the servers he wants to con-nect. Clients for MetaFrame XPaare available for all major operat-ing systems like Windows, Linux,MacOS and IRIX. The client isable to map local drives and print-ers to the server.

The Terminal Server Farmwas widely accepted by users.Not only Linux users, but alsomany Windows users installedthe client and started to use soft-ware on the server. This pro-foundly reduced software instal-lation requests on user machines.As a second effect software costswhere also reduced. Expensivescientific software was placed onthe server farm and could be usedthere by all interested users. Alsonon-standard graphics software,which is only needed to view orconvert files received from co-workers, is restricted to terminalserver usage.

Although the Terminal ServerFarm was a huge success someproblems occurred during theroll-out phase. Not all softwarewas able to run on MetaFrameservers.

In conclusion the roll-out ofthe Terminal Server Farm was asuccess. Though the break evenis not yet reached, user numbersare increasing while support re-quests are decreasing. By pro-viding a standard TeX environ-ment to all users, the TerminalServers even helped to producethis report with less effort thanlast year.


Realization of InstrumentSubnetworks

By providing an individual broad-cast domain to each instrument,interactions between instrumentcontrol software should be keptto a minimum. The networksare designed in a way, that theycan operate independently fromthe FRM-II network, in case offailure. Each instrument net-works contains a switch and theinstrument control hardware re-quired. In addition it is highlyrecommended to install an instru-ment server. This server pro-vides basic services like DNS,DHCP or File Sharing (NFS orCIFS). Acquired data will betemporarily stored on the instru-ment server, a failure of the cen-tral FRM-II file server will there-fore not influence instrument op-eration. The same holds truefor DNS and DHCP, in additionthe local servers help to reducenetwork traffic by keeping localtraffic local. If no additionaltasks should be performed by thisserver, a mid-class PC (P-III 1GHz or above) is sufficient.

The Taco-Box LinuxDistribution

The Taco-Box Linux distributionfor embedded instrument controldevices has been renewed thisyear. The idea is to place smalldevices running Taco next to thehardware to be controlled. Oneapplication for this is to provideEthernet connection for deviceswith just a serial port. OurTaco-Box Linux Distribution isderived from SuSE 7.2. To-gether with Taco it fits on 6

disks, occupying 10 to 12 MBafter installation. So far wehave used the Taco_box distribu-tion on JumpTec and ICP Wafercomputers. The first disk con-tains the boot image, the nexttwo disks Tacobox A and B con-tain the major operating system,disk 4 contains SSH and SSL,disk 5 the kernel and disk 6 theTaco system itself. Installationcould either be done with mon-itor and keyboard, but the lat-est release also supports a serialconsole for installation. Afterinstallation the system could beaccessed either by keyboard, se-rial console or remotely via ssh.The Taco-Box Distribution sup-ports NE1000, NE2000, CirrusLogic CS8900 and CS8920 net-work cards, multi-port serial in-terface cards and the PC104 bus.

Details about the Taco-BoxLinux distribution and disk im-ages can be found on the FRM-IIwebsite.

Planing of a centralDocument- andWorkflow-ManagementSystem

In conjunction with the planingand construction of FRM-II largeamounts of paper have been cre-ated and must be stored. Further-more paper files might get mis-placed or damaged. The ideais to store all legal required in-formation about the FRM-II andits operation in an electronic stor-age. Such Document Manage-ment Systems (DMS) are alreadyin use and become more popularwith eGovernment projects. Suchsystems may not only be usedto store documents but to also

define a work-flow for repeatingtasks.

As a pilot study the first stepis to bring a system for electronicdistribution and storage of snailmail and for electronic procure-ment into operation. Require-ments for such a system are, thatit is possible to use it from Linuxand Windows clients and thatthe server should also run underLinux. We decided to use theSoftware Fabasoft, which is al-ready in use at major AustrianGovernment offices.

Incoming mail is first scannedand then electronically for-warded to the recipient. Thescanned document can be putinto an electronic file or passedon. Order forms can be filled inon the system and electronicallypassed on for approval. The de-fined workflow then processesthe order and passes it on to allpersons who need to approve it.Finally the approved order comesback to the issuing user. The userthen can implement the order andalso passes on the approved formto the administration. When theadministration receives the billthe ordering user has to approvethe bill and the administrationpays it. The whole process isdone electronically. If the sup-plier needs the order on paper, itis possible to create a printout ofa standard order form from thesystem, but the rest of the order-ing process is filed electronically.

A pilot system for 50 usersis installed and training of pilotusers takes place in the moment.The roll-out of the order process-ing workflow will take place atthe beginning of next year. Af-ter 6 month the process will be


evaluated and a decision on fur-ther steps taken.

[1] Zeilenga K., Chu H., Masarati P. OpenLDAP.http://www.openldap.org/.[2] Tridgell A., et al. The Samba Project.http://www.samba.org/.[3] Project F. S. The FreeRADIUS Server.http://www.freeradius.org/.[4] Bricart C., Link R. AMaViS – A Mail Virus Scanner.http://www.amavis.org/.[5] Hrustica V. Sophie.http://www.vanja.com/tools/sophie/.[6] Dinter T. V., Findlay D., Hughes C., Mason J., Quilan D., Sergeant M., Stretz M. S. SpamAssasin.

http://au.spamassassin.org/index.html.[7] Castello R., et al. Squirrel Mail – Webmail for Nuts.http://www.squirrelmail.org/.

1.4 The Neutron Guides at FRM II

Ch. Schanzer1, E. Kahle2, G. L. Borchert 2, D. Hohl 3, S. Semecky4

1 Physikdepartment E 21, TU München2 ZWE FRM-II, TU München3 D. Hohl, Technische Dienstleistungen, München4 S. Semecky, Kommunikationstechnik, München

During the year 2003 the con-struction of the neutron guides(NG) at FRM II was accom-plished. It comprises the follow-ing tasks:

1) The adjustment of theNG was performed by meansof theodolites reaching high

Figure 1.6: The tunnel wall, seenfrom the casemate side after com-pletion. The vacuum tubes stick-ing out of the steel plates are con-nected with the high speed shut-ters. At their exit the vacuumtight NG follow, which are sup-ported by the steel bars (first oneis brown).

mechanical tolerances. Maxi-mum displacement of two conse-quent glass elements is less than10µm, at complex positions, asjunctions, less than 20µm, angu-lar mismatch less than 15arcs andless than 45arcs in horizontal andvertical direction, respectively.

2) In section 0 and section1 of the FRM II the NG areenclosed in stainless steel tubesequipped with conventional vac-uum fittings. These sections havebeen checked for vacuum tight-ness of less than 2· 10−2 mbar.In section 2 of FRM II (NG halland casemate) the glass NG haveto be vacuum tight, except NL2b. Therefore all consequentglass tube elements were gluedtogether. The final element wasequipped with an Al window thatwithstand the vacuum pressureand is transparent for neutrons.

3) All NG in the NG hall have

been jacketed by 7cm thick leadplates.

4) The three openings in thetunnel wall where the NG NL1to NL6 pass through have beenclosed. To reach the maximumshielding a sandwiched structure

Figure 1.7:Closing the 80cm thickcasemate wall towards the NGhall. Between the lead jacketedNG and the high density concreteblocks (brown) are visible thelight-grey steel containers, filledwith concrete, which are preciselyfitted into the remaining gaps. Onthe left side the steel plates for fi-nal closure can be seen.


of materials was used. Startingfrom the casemate side, we in-troduced one layer of 10cm lead,one layer of borated PE, 16cmthick. The remaining 155cm ofdepth were filled with Röbalithtype 4 bricks (density 3.2g/cm3).As the NG pass at skew direc-tions through the openings thebricks were cut so that they fitprecisely in the gaps. All bricksare arranged in a way so that nothrough-running slits occur. Re-maining slits at the borders wereclosed with lead wool. From out-side the openings were coveredwith steel plates (see Fig.1.6).

5) The openings in the case-mate wall were closed with highdensity (4.5g/cm3) concrete units.To scope for the skew directionsof the individual NG stainlesssteel containers were producedthat fit precisely in the prismaticgaps (see Fig.1.7). They werefilled with the high density con-crete and inserted into the gaps.Remaining slits at the borderswere closed with steel

Figure 1.8:the completed casematewall. The individual NG run-ning through the casemate passthe wall towards the NG hall.

Figure 1.9:The accomplished wallof the NG hall. On the final steelcover several shock sensors of thesafety system are visible. All NGare jacketed by the (blue) leadplates.

plates or lead wool. Finally theopenings were covered on bothsides with steel plates. In Fig.1.8and Fig. 1.9 the completed wallbetween casemate and NG hall isshown. A final overview of theNG hall is depicted in Fig.1.10.

Figure 1.10:A general view of the NG hall. The completed NG run to theirinstruments or end at the position where in next future the correspondinginstruments will be built.

6) The high speed shutters ofNL1 to NL6 have been installedcompletely and integrated intothe central control system. Ina first test it was observed thatthe accuracy of the used vacuumsensors was not sufficient for areliable operation. After somesystematic tests they all were re-placed by another high accuracytype. Now the complete shut-ter system works in the expectedway. For three NG we installedthe corresponding instrumentalshutters.

Finally all tasks have beencompleted successfully so thatthe instruments can make full useof the NG facility, provided neu-trons are available.


1.5 Progress Report of the Sputtering Device for Supermirrors

C. Breunig1, Prof. Dr. G. Borchert 1, A. Urban 1


Production ofSupermirrors

For conducting the thermal neu-trons from the source to variousexperiments the research-reactoris equipped with 437 meters ofneutron guides, produced by twocompanies, partly in cooperationwith the sputtering-team of theFRM II. As this team operates aself made sputtering device, partof the supermirrors i.e. glass-plates with thin multilayers con-sisting of increasing alternatinglayers of titanium an nickel ina magnetron sputtering processwere produced there. Finally theglass-plates were glued togetherto build a rectangular vacuumtight neutron guide. To checkthe performance our supermir-rors are measured at the neutronbeam of BENSC in Berlin. Af-ter concluding 75 sputtering pro-cesses which means an amountof 11,7 m2 super mirrors of qual-ity 2.0 and 1.8 m2 of quality 2.5from January up to the end ofFebruary in 2003, the sputteringdevice was dismantled.

Rebuilding the facility

Apart from the fact that thelaboratory was renovated andjoint with a neighboring roomsome important modifications atthe sputtering device itself werestarted. To fill the conditionsof getting optimal atmosphereand clean circumstances the slide

vane rotary vacuum pumps wereinstalled in the basement un-der the laboratory for keepingthe atmosphere oil-free. Ad-ditionally there was installed acomplete cleanroom cabin (pu-rity class 1000) with a directair lock to the sputtering de-vice for encamping, preparingand cleaning the glass before the

Figure 1.11:Sputtering facility

sputtering process.Apart from these modifica-

tions of the surrounding area ofthe sputtering plant some newfeatures were mounted directly atthe installation itself. The mostobvious innovation is a thirdvacuum chamber fitted directlyto the main chamber where thesputtering-procedure takes place.


In this sluice the cleaned glass-plates will be prepared for thesubsequent process by using ainfrared substrate-heating in or-der to detract last water odd-ments. Furthermore for guar-anteeing a dirt-free surface, theglass is cleaned by a fluoresc-ing plasma to debar last impuri-

ties. One more positive effect ofthis new sluice is, that the pro-cess itself is accelerated becausepumping times can be reducedand parallel procedures are pos-sible. Another innovation wasdone inside the main vacuum ves-sel, the new drive train consistingof a link-age for facilitate consis-

tent speed and ensure the drive-train-unit maintenance-free. Forthe whole process control a newLabview system is installed, en-abling a fast and simple commu-nication with the device.

With the beginning of theyear 2004 the production will berestarted with improved quality.

1.6 Sample Environment

J. Peters1, H. Kolb 1, A. Schmidt1, A. Pscheidt1, J. Wenzlaff1


In 2003 the main activitieswere focused on further develop-ment of our closed cycle refriger-ators and high temperature equip-ment.

Closed Cycle Cryostat

Extensive tests were done on ournew sample tube closed cycle re-frigerator CCR2 resulting in animproved performance. Featuresare the compact design and easyhandling. The initial cool downtime from room temperature to3.5 K now is about 4 hours.Three additional cryostat’s willbe ready for use early 2004.Sample holders allowing temper-atures from 3.5 K up to 700 Kare available. Sample cool downtimes are about 1 hour depend-ing on initial temperatures. In co-operation with the Walther Meiss-ner Institute, Garching,3He/4Hedilution inserts are under con-struction. Appropriate gas han-dling systems are being mountedactually. In co-operation with theFRM II software team, an auto-matic cool down procedure wassuccessfully tested.

Magnet

After completing the design re-view our cryogen free supercon-ducting split coil magnet is un-der construction at ACCEL In-struments GmbH. The magnetwill provide a vertical symmet-ric magnetic flux of 10 T, a splitof 30 mm and a room tempera-ture bore of 100 mm. Delivery isplaned early 2004.

High Temperatures

In co-operation with Paul Scher-rer Institute (PSI), a FRM-IItype high temperature furnacebuilt for the structure powderdiffractometer (SPODI) was suc-cessfully tested with neutronsin April 2003. SPODI instru-ment responsible R. Gilles andco-workers accomplished SANSmeasurements on Ni-base super-alloys at temperatures up to 1700K.

An IR-radiation furnace allow-ing temperatures up to 1700K, providing a 40 mm beamwindow and a horizontal ac-cess of 360° is under develop-

ment. The furnace can be evac-uated or operated using diverseatmospheres. Heating is donevia four commercially available

Figure 1.12:Top loading Closed Cy-cle Cryostat CCR2


reflectors equipped with 150 Whalogen lamps. To achieve high-est temperatures the sample sizeis restricted to the the filament di-mensions (3 x 5 mm). First testsare very promising.

High Pressure

A co-operation with the Bay-erisches Geoinstitut Bayreuth hasbeen initiated for the develope-ment of a diamond anvil cell suit-able for neutron scattering exper-iments.

Figure 1.13:IR-furnace

1.7 Software Group

J. Krüger 1, S. Praßl1, H. Gilde 1, S. Roth1, U. Graseman1, S. Huber1, A. Fuchs1, M. Salfer 1,S. Kreyer1, J. Beckmann1


The main activities of thesoftware group of the FRM-IIin 2003 were on the followingfields:

• Improvement of the basicTACO system in close co-operation with the ESRF

• Improvement of currentTACO servers

• Development of a databasedriven system for creationand management of TACOdevices and servers

• Enhancement of the TACOLabView server

• Enhancement of the

"TACO-box" in collabo-ration with the IT/Networkservice group

• Recoding of the open-DaVE software


TACO basic system

In 2003 software groups of theESRF[1] and FRM-II decidedto put the TACO system onSourceforge[2]. The main ad-vantage is to get an infrastruc-ture for developing and improv-ing the TACO system. In view oferror handling and future exten-sion. The first step was to mergethe code bases of the ESRFand FRM-II into a common one.Into this merge a new designeddatabase server entered, whichmay use the "old" GDBM[3]as well as the "new" mySQL[4]database. The mySQL databasegives the user much more flexibil-ity for changing resources.

In addition to the current sys-tems we adopted the code to runalso on a FreeBSD system whichallows to use the TACO systemalso on BSD based systems likeMac OS X.

TACO servers

During the last year we improvedand stabilised the existing TACObasic servers. Some new basicservers were written, especiallyservers which provide the com-munication through several bussystems and several communica-tion protocols.

The following servers belongto this group:

RS232

With the help of this server onemay communicate directly witha connected device. That wayone may send a request or com-mand to the device and get theanswer from it. The communi-

cation is not restricted to ASCIIcommands, it can also be used forbinary communication protocols.An additional feature is the possi-bility to send series of communi-cation requests, without interrup-tion.

Server for ModBus ProtocolGöttingen

Server for GPIBThis new server was developed

for the communication over theGPIB interface. The first ver-sion of this server was based onthe developments of the Linux-GPIB[5] group on Sourceforge,which restricted the possible se-lection of GPIB cards. However,this library supports most of thestandard cards.

In a next step the serverwas extended to the NationalInstrument High-PerformanceEthernet-to-GPIB-Controller NIGPIB-ENET/100. This con-troller will be used by the PUMAexperiment. Due to the librarycompatibility of the free imple-mentation to the NI library it wasonly necessary to bind the NIlibrary to get a running system.

This server has the same set ofcommands as the RS232 server.Thus one may change the com-munication path without chang-ing the software (tested on theLakeShore 340 controller, withRS232 and GPIB communicationinterface).

Server for Profibus DPBased on the server developed

at the ZEL at the FZ Jülich[6]we have completely rewritten theserver for the Profibus DP. Thereason for the rewrite was to ex-tend the server by some new com-mands for reading and writing

• bits

• words (short int’s inC/C++)

• double words (long’s inC/C++)

• floats• double floats

directly to the Profibus DP.The advantage of direct writingor reading these types directly isto avoid the conversion of thedata on the client side resulting insimpler use of the Profibus DP.

As the old server uses the OIC(Objects In C) model, and allservers at the FRM-II are writtenin C++ we decided to rewrite thisserver.

In addition to the communi-cation TACO servers we devel-oped some servers for more spe-cialised tasks.

Temperature controller like Eu-rotherm 2404 and LakeShore340:

Both servers provide the sameset of commands so it will bepossible to use the same clientfor both servers. This allows theuser to use parts of software de-veloped by other users.

Vacuum gauge Leybold IT90and and Vacuum gauge controllerITR23

Both devices provide the sameset of TACO commands. Thusthe client software does not needto support both. The device mayalso be used as analogue input de-vices. This is a standard deviceclient class for all devices whichgive analogue inputs to the exper-iment control software.

Turbo molecular pump con-troller NT20

This device only gives some in-formation about the status of thepump controller. The controller


does not accept any commandsfrom the serial line.

Motor control of the HeiDi ex-periment

The HeiDi experiment uses aSIEMENS PLC S7 for control-ling the motors and encoders ofthe sample table and the detector.The S7 is connected to the com-puter by a Profibus DP interface.This server exports right now 4motor and 4 encoder devices toaccess all four axes. It will be ex-tended by a device to control allaxes simultaneously.

TACOdevel

This project is a database drivensystem for developing and man-aging the TACO devices andservers. With the increasingnumber of TACO devices andservers one gets problems withthe TACO commands of theservers.

TACO implies the followingrules:

• Devices belong to classeswith the same set of com-mands.

• Commands are only num-bers

• Servers export devices

The problem of this philoso-phy is that there is no way tocheck the rules. Therefore we usea mySQL database to store thecommands of all devices as wellas the information which servermay export what devices. Fromthis database one may create acode frame into which the cur-rent implementation of the com-mands of the devices has to becoded. This framework allows usto speed up the development of

new TACO servers because thecoding of the setup of a server istaken by the framework.

Another advantage ofTACOdevel is the consistencyof all TACO clients. The num-ber of client classes is drasticallyreduced because only for onetype of device a client class ex-ists. Usually the control softwareneeds mainly the following typesof devices:

• analogue outputslike motor positions, volt-ages, temperatures.

• analogue inputslike encoder positions,voltages, temperatures.

• digital inputslike counter presets, timerpresets, switches.

• digital outputslike counter values, limitswitches, timer values.

Each device, additionally to itsbasic functionality, needs somecommands for setup etc. There-fore we divide these basic groupsinto subgroups, summarizing de-vices with same functionality liketemperature controllers, motors,or temperature sensors. Theseclasses inherit the functionalityof their base classes.

Thus, the first step of devel-oping a TACO device for a newhardware reduces to look for asimilar device in the database andto integrate this device into thenew server. After this, the basiccode has to be generated from thedatabase. Afterwards the real im-plementation of the server devicestarts. Usually the client is readythen.

TACO binding toLabView

Based on the development byAndy Götz from the ESRF we de-veloped a new TACO server to in-terface LabView. We looked fora possibility to integrate this pow-erful control software. Becausesome commercial products, espe-cially sample environments likehigh field magnets, are controlledby LabView programs. LabViewis shipped with some powerfulmechanisms to execute external(native) codes. One of thesemechanisms allows developers toinvoke external library functioncalls from within LabView. Thisport was used at the ESRF, andwe used it to access the LabViewcontrols and indicators. With ourextension it is possible to inte-grate LabView VI’s (virtual in-struments) into a LabView TACOserver.

The access to LabView withthe help of our solution is kindof generic and should work withmost VI’s.

Additionally we tested the net-work transparency of LabView(the LabView Control may runon a different computer than theTACO LabView server).

TACO box

"TACO box" is a small floppydisc based linux distribution de-rived from a SuSE 7.2 distribu-tion, which can be installed onan embedded PC with reducedfunctionality. It provides onlythose features necessary to runa TACO system including thedatabase and some simple TACOservers like RS232.


The main application of suchservers is the TACO binding ofsample environments, like hightemperature ovens, or cryostats.With the help of such embeddedcontrollers add these sample en-vironments can be added into theexperiment control software veryeasily. The hardware is addedby plugging in a network cable,and the software by importing theTACO devices.

Reengineering ofopenDaVE

In the first half of the year theold openDaVE was rewritten byone of its creators, U. Graseman.It was shown, that some of thefirst implementation details wereclumsy implemented. The sys-tem contained some restrictionswhich were not very helpful inpractice.

These restrictions are removedin the new implementation.

It changes the design to a lessrestrictive system. The data flow-ing between the modules may bedata in the old sense, commands,events or whatever. So the sys-tem became more flexible thanthe old one.

[1] http://www.esrf.fr.[2] http://taco.sourceforge.net.[3] http://www.gnu.org/software/gdbm.[4] http://www.mysql.org.[5] http://linux-gpib.sourceforge.net.[6] http://www.fz-juelich.de/zel.

1.8 NICOS – The ease of getting TACO servers

T. Unruh 1, C. Geisler1


Since the development of thefirst version of the networkbased instrument control systemNICOS in 2001 it is advancedto an accepted frontend for theTACO system at FRM–II. Be-sides the NICOS client espe-cially the NICOSMethods whichhave been described in the FRM–II annual report 2002 are in-tensely used. NICOS is currentlyused at 4 instruments at FRM–II: The polarised cold neutronthree–axis spectrometer PANDA,the thermal three–axis spectrome-ter PUMA, the MIRA reflectome-ter for very cold neutrons andthe time–of–flight spectrometerTOFTOF.

In 2003 some new featureshave been added to NICOS. The

most important add–on whichwill be described in the follow-ing is a generic TACO server(nicmTACO) which allows to ex-port NICOS methods as TACOservers at the push of a button.

These TACO servers are fullyfunctional and are running out-side the nicos environment. Theuser has just to set two param-eters in the configuration edi-tor GUI. No resource files areneeded, no special konwledgeof the TACO internals are re-quired to write a class that canbe exported as a TACO de-vice. The nicmTACO systemmainly consists of three com-ponents: The nicmServer andnicmClient classes and a Server-Factory script. The fundamen-

tal class nicmServer creates andembeds an instance of a nicosmethod, reads its configurationand provides access to its publiccommands. The ServerFactoryscript instantiates one or morenicmServers and exports themas device on the TACO system.The nicmClient class instancecan be used to communicate witha given TACO server; further-more, if that specific device is notyet exported via TACO, it takescare during its init process, thatthe neccessary server is startedautomatically.

The only visible differencefor the user is as mentionedbefore, that he has to changethe server parameter of his de-vice from ’nicos’ to ’TACO’ and


give a valid TACO name forthat particular device in the form〈domain〉/〈family〉/〈member〉 us-ing the configuration editor. Theresulting client instance is equiv-alent to a pure nicos method withrespect to the use of public com-mands. Some additional func-tionality is available, which isalso helpful for development anddebugging purposes as e.g. an au-tomatic server–restart procedureand a logging mechanism.

The current state of develop-ment is, that any derivative of theXable or HWGeneric classes canbe exported as a TACO server.

Grouping via XOGroup workstoo. The controller derivativeis currently under ’heavy testing’and is most likely available in astable version in spring of 2004.

Another remarkable feature ofnicmTACO is, that there is nomore need to write resource filesto get the server parameters intothe TACO database. The nicm-Server init process creates alldatabase information on–the–flyout of the configuration, whichthe user generates using the con-figuration editor. You can run thenicmTACO system nearly with-out any restrictions outside from

NICOS. Only a python installa-tion (version> 2.1.0) is needed:Just import the nicmClient orServerFactory module, instanti-ate a nicmClient class with thekey of the configuration of yourdevice and run the start() com-mand with that key as parameter,respectively, and your server isup! The only drawback of usingnicmTACO outside the NICOScontext is that some functionalityrequiring direct access to a run-ning NICOS daemon may (in thecurrent state of development) notwork.

1.9 Radiation Protection

H. Zeising1


Research Reactor FRM-II

On the day of 02.05.2003 the TU-München received the 3rd partiallicense according to §7 AtomicEnergy Act (AtG) from theBavarian State Ministry for StateDevelopment and Environmentfor the operation of the high fluxneutron source Munich (FRM-II). The license also encompassesthe handling of nuclear fuel andother radioactive substances andhas the stipulation of immediateexecution.

Prior to this an intensive ap-praisal was undertaken by theFederal Ministry for the Envi-ronment, Nature Conservationand Nuclear Safety (BMU), whobrought in not only the ReactorSafety and Radiation ProtectionCommissions, but also externaloutside experts. The appraisals

were concluded with positive re-sults.

Due to the direct execution,directly after reception of the

Figure 1.14:The inspection of the screw cap of a cask filled with heavywater.

license for operation and han-dling the delivery of the heavywater (D2O), which is the moder-ation medium needed for the


operation of the reactor, and thedelivery of the first two fuel rodsand the two converter plates forthe beam tube converter facility(SKA) were initiated. In addi-tion, the primary start-up neutronsource was installed in the moder-ator tank.

Delivery and filling in ofthe heavy water (D2O)

The heavy water moderationmedium bought for FRM-II,which was kept in storage inCadarache in France, had alreadybeen used in a French research re-actor, after which it was cleanedand detritiated. It still has a tri-tium activity of about 2× 1010

Bq/l.The amount of heavy wa-

ter needed for FRM-II is about20 m3 and was transported toGarching in two charges, packedfor transport in 105 stainless steelcasks with a volume of 200 l.

The first delivery with 52 caskstook place on 05.06.2003 andproceeded smoothly. On thesecond transport with 53 caskson 13.06.2003, two leaks weredetected on the screw caps oftwo casks during the radiationsafety‘s incoming componentsinspection. Since the criticalvalue for the contamination of ra-dioactive transport was exceeded,the leaks had to be reportedto the atomic controlling institu-tion. The amount of heavy wa-ter which leaked out was about 5ml. This small amount of leakedwater drew an extensive control-and review procedure after it inorder to found the cause of theleakage. The procedure was not

yet concluded at the end of theyear 2003.

In the time period from thebeginning of October until themiddle of November 2003 therequired amount of heavy waterwas filled from the casks intothe D2O system of FRM-II. Fromthe viewpoint of radiation pro-tection extensive safety precau-tions were necessary. Tritium isa low energy beta particle emitterwhich causes practically no exter-nal radioactive exposition. How-ever, when handling open tritium,

Figure 1.15:Pulling out of a fuel rod from a transportation container in thetruck lock of FRM-II.

precautions must be taken inorder to avoid dose-relevant in-corporations of the personnel.One must pay attention that thetritium concentration in the airof the room where the casks areopened and filled into the systemstays at a low level.

After completion of each workcampaign the participating per-sons had to deliver urine sam-ples which were analysed bythe official evaluating Office, the"Messstelle für Radiotoxikolo-gie" at the Bavarian Environ-


mental Protection Agency inKulmbach, in order to find thetritium activity in the body andcalculate the resultant personaldoses. The internal radiationdose was not higher for any sin-gle person than a few microsiev-erts. The collective dose for theentire filling in of the heavy wa-ter was about 60 microsieverts.There were no unplanned eventsduring the filling.

One can say that the fillingin of the heavy water proceededvery satisfactorily.

Delivery and storage ofthe fuel rods andconverter plates

The fuel rods and the converterplates for the beam tube con-verter facility (SKA) are manu-factured in France. The deliv-ery from France proceeded understrict conditions using a so-calledsafety vehicle (SIFA) to FRM-IIin Garching.

The first two fuel rods anda converter plate were deliveredon 10.07.2003. Due to reasonspertaining to technical authoriza-tion of transport, the second con-verter plate was delivered extraon 23.07.2003.

The fuel rods as well as theconverter plates contain uraniumwith an Uranium-235 enrichmentof about 93 % by weight, so-called HEU (highly enriched ura-nium).

The delivery, unloading andthe storage of the two fuel rodsand the two converter plates inthe fuel element repository wasplanned well and proceeded with-out any problems.

Special radiation protectionmeasures were not necessary dur-ing the unloading. The fuelrods and the converter plateswith the firmly integrated ura-nium showed only a minor doserate. The dose rate was onlyabout 20 microsievert per hour.

Just one day later on11.07.2003 a first inspection ofthe stored nuclear fuel was con-ducted by the IAEA and Euratom.It took place without any com-plaints.

The installation of theprimary start up neutronsource, Cf-252

In order to start up the reactorand to permit the smooth regu-lation using the control rod, onehas to have a certain neutronflux at one´s disposal even be-fore criticality is reached. Thisis achieved through use of a Cf-252 source. The Cf-252 decaysto a small part through sponta-neous fission and sends out neu-trons at the same time. The pri-mary start up source for FRM-IIhas an activity of 1.2× 1010 Bqwith a neutron emission rate of1.4× 109 per second. The neu-tron source was manufactured inEngland, delivered to Garchingon 25.04.2003 and on 01.12.2003it was installed in the moderatortank of FRM-II.

Because of the relatively shorthalf-life of Cf252 of 2.645 yearsit is essential to have a furthersecondary start up source avail-able for later start ups. This ismade possible through use of asecondary start up source whichis made up of antimony and beryl-

lium. Through activation of theantimony (Sb-123) during the op-eration of the reactor to Sb-124and the resulting high energygamma radiation through a (γ,n)reaction of the Be-9 to Be-8 orrather, two alpha particles, the re-quired neutrons are generated.

Instruments andExperiments

In the year 2003 further propos-als regarding the design of in-struments and experiments weretalked through and discussedwith the submitters.

As a result, the assembly of theapproved instruments proceededquickly.

Information brochure “Allaround safe!”

In July 2003 an informationbrochure was completed, re-quired according to paragraph§53 of the Radiological Protec-tion Ordinance, which informsthe public about precautions tofight damage at the onset of sig-nificant safety-related incidentsat FRM-II.

The brochure with the title“All around safe!” was sent viabulk mail to all households ofthe communities within a 2 kmperimeter to FRM-II.

Other interested parties canobtain the brochure at the pub-lic relations office of FRM-II at all times or online fromhttp://www.frm2.tum.de/publikationen.

24 2 Diffractometers

2 Diffractometers

2.1 BIODIFF – Monochromatic Single Crystal Diffractometer forBiological Macromolecules

A. Ostermann2, F.G. Parak1

1 Physik Department E17, TU München2 ZWE FRM-II, TU München

It is planned to install amonochromatic single crystaldiffractometer at the thermalbeam tube SR5a at the FRM-II.Monte-Carlo simulations (Mc-Stas) have been carried out tooptimize all components of theinstrument. This diffractometer

Neutron guide

Velocity selector

Mobile monochromator shielding

Sample goniometer

Monochromator

Neutron imaging plate

a

b

Figure 2.1:Schematic view of theplaned diffractometer. (a) Topview. Only a part of the neu-tron guide is shown. (b) 3-dimensional view.

Monte-Carlo simulations (Mc-Stas) have been carried out tooptimize all components of theinstrument. This diffractometerwill be dedicated to the structuredetermination of proteins andother biological macromoleculesin the temperature range between20K and 300K.

Hydrogen atoms play an im-portant role in biological macro-molecules. They mediate hy-drogen bridges and take part innon-bonding interactions such aselectrostatic and van der Waalsforces. They are essential forthe stabilization of the defined3-dimensional structure of pro-teins and thus contribute to thecomplex energy landscape of thismolecules. Hydrogen atomsin polarized bonds play criti-cal roles in enzymatic catalysis.They are involved in hydrogenbonds in the substrate bindingand are essential in proton trans-fer reactions during the cataly-sis. The knowledge about aminoacid protonation states in theactive center of proteins is of-ten crucial for an understandingof the molecular reaction mecha-nism. Another point of interestis the hydration structure around

proteins. This hydration struc-ture has a great influence on thedynamic properties of proteins[1,2].

The planned diffractometershould allow the determinationof hydrogen atom positions andmean square displacements inproteins. For a detailed dataanalysis it is important thathigh resolution data sets canbe collected[3, 4]. The planneddiffractometer is specially de-signed for the data collection ofcrystals with large units cells (≥100Å·100Å·100Å). Due to thelarge unit cells of protein crys-tals the total diffracted intensityis distributed over a large num-ber of Bragg reflections and istherefore weak. For this reason alow background and a small hori-zontal and vertical divergence isessential for this instrument.

Instrument design

To optimize all componentsMonte-Carlo simulations (Mc-Stas) of the whole beam path,including the scattering processat the sample crystal, have beencarried out. An overview over

2 Diffractometers 25

the main components of the in-strument is shown in figure2.1.

Neutron guide

The geometric design and coat-ing of the neutron guide aredetermined by the monochroma-tor characteristic and thereforeby the target value of the diver-gence at the sample position. Toachieve a low background, the in-strument will be installed at a 8.7m neutron guide. The plannedneutron guide consists of 5 chan-nels (channel width = 5.2 mm)and is slightly curved (R = 1800m) to avoid a direct sight to thebeam tube entrance. This willsuppress the background of fastneutrons andγ-radiation. Due tothe small divergence at the sam-ple position a coating with58Niis sufficient. The guide will beslightly vertical focussing withan entrance height of 96 mm andan exit height of 57.5 mm. Thetotal guide width is 30 mm.

Velocity selector

The next component in the beampath is a lamella velocity se-lector (Dornier). This selectoris used as a higher order fil-ter for the monochromator (py-rolytic graphite) and will allowa continuous wavelength choicewithin the wavelength range of in-terest (λ = 2.0− 4.0Å). The ve-locity selector combines a veryefficient energy cut–off (spectralband width∆λ/λ = 35%) witha high transmission of 87.5% –85%. Due to the energy cut-offthe selector is also a essentialcomponent for the background re-duction.

Wavelenght(Å)

Hor. diver-gence FWHM(°)

Vert. diver-gence FWHM(°)

∆λ/λ FWHM(°)

2.0 0.71 0.72 2.12.75 0.68 0.83 1.83.5 0.64 0.90 1.34.0 0.62 0.90 1.1

Table 2.1:Results of Monte-Carlo simulations for the divergence and wave-length distribution at the sample position (2mm x 2mm).

Monochromator unit

The diffractometer will be able tooperate in the wavelength rangeof λ = 2.0− 4.0Å. However, inits standard configuration, thediffractometer use will neutronsof approx. 2.75Å to cover alarge area of the reciprocal space(dmin ' 1.45Å). As monochro-mator a flat highly oriented py-rolytic graphite monochromator(PG(002), mosaicity = 0.4°–0.5°)will be used. Pyrolytic graphiteshows in the wavelength rangeof 2.0 - 4.0Å a high reflectiv-ity of 63% – 80%. Since thedistance between the monochro-mator and the sample positionis rather small (≤ 1.7 m) no fo-cussing geometry will be usedto keep the divergence withinthe desired range. The planedmonochromator has a width of75 mm with a height of 35 mm.The thickness is 2 mm. This flatmonochromator is a easy and in-expensive solution. Furthermorethe background will be reducedsince apart the monochromatorthere is nearly no supporting ma-terial in the direct beam.

For the cylindrical monochro-mator shielding different types ofheavy concretes with densities

up to 6.3 g/cm3 will be used. Theouter diameter will be 1940 mm.The shielding will allow a contin-uous usage of of 2ΘM-angles be-tween 30°and 75°. A flight tubeconsisting ofB4C/resin and leadbetween the monochromator andthe diffractometer will be used asa first collimation unit.

Diffractometer

For the detector of the diffrac-tometer a neutron imaging plate(NIP) will be used. Due tothe large dynamic range andthe very high spatial resolu-tion (≤200µm) neutron imag-ing plates are the best choicefor diffraction experiments, espe-cially in the case of large unitcells. The neutron imaging platesare mechanical flexible and thuscan be easily adapted for the de-sired geometry. Neutron imagingplates are already successfullyused for the diffractometers BIX-3/4 (reactor JRR3M, Japan) andLADI/VIVALDI (Institut Laue-Langevin, France).

To cover a large solid anglerange the NIP will have a uprightcylindrical geometry. The sam-ple is placed in the middle of thecylinder. The vertical cylindri-cal design allows a high flexibil-


ity concerning the sample envi-ronment. For the NIP-cylinder ra-dius one has to consider the beamdivergence at the sample position(compare table 1). Monte-Carlosimulations of the planed instru-ment show that a radius of 230mm is the best choice. The to-tal covered horizontal solid an-gle is±145°. The total NIP-areais 1160 mm (width) by 550 mm(height) and will be realized intwo pieces.

The NIP readout device is com-posed of a semiconductor laserand a photomultiplier. Duringthe readout procedure the NIP-cylinder is rotated at high speedwhile the reader moves from oneside to the other. Finally, any re-maining signal is erased using ahalogen lamp. The time for read-out and erasing will last about5 minutes. To shield the NIPagainst a potential neutron- andγ-background from the reactor hallthe diffractometer is covered bycylindrical shielding ofB4C/resinand lead.

-50-100-150 50 100 1500

-2 -1 0 1 2[cm]

2

4

6

8

10

2

4

6

8

10

-2 -1 0 1 2[cm]

2

4

6

8

10

-2 -1 0 1 2[cm]

2

4

6

8

10

-2 -1 0 1 2[cm]

1.5A 2.0A 1.5A2.0A

[degree]

λ=2.75Α

[cm

]

[cm

]

[cm

]

[cm

]

Figure 2.2:Monte-Carlo simulation of Bragg reflections from a sample crys-tal. Only the upper half of the neutron imaging plate is shown. De-tailed views are shown for a resolution of d=2.0Å and d=1.5Å. Unit cell:100Å·100Å·100Å, α = β = γ = 90°; crystal mosaicity = 0.1°; neutronimaging plate spatial resolution: 0.4mm x 0.4mm.

Results of Monte-Carlo simu-lations of the scattering processat a sample crystal with unit celldimensions of 100Å·100Å·100Åare shown in Fig. 2.2. With awavelength of 2.75Å it is possi-ble to measure crystals with

such a large unit. With a wave-length > 3.0Å it will even bepossible to measure crystals withlarger unit cells. A big advantageof this instrument is the possibil-ity to adapt the wavelength to theunit cell of the sample crystal.

[1] Bon C., Lehmann M. S., Wilkinson C.Acta Cryst. D, 55, (1999), 978–987.[2] Chatake T., Ostermann A., Kurihara K., Parak F. G., Niimura N.Proteins, 50, (2003), 516–523.[3] Ostermann A., Tanaka T., Engler N., Niimura N., Parak F. G.Biophys. Chem., 95, (2002), 183–193.[4] Engler N., Ostermann A., Niimura N., Parak F. G.Proc. Natl. Acad. Sci. USA, 100, (2003),

10243–10248.

2.2 HeiDi – Single Crystal Diffractometer at the Hot Source

M. Meven1, F. Hibsch1, G. Heger2

1 ZWE FRM-II, TU München2 RWTH Aachen, Institut für Kristallographie

The single crystal diffractome-ter HEiDi uses beam line SR-9 inthe experimental hall of the FRM-II. This beam tube is the only

one that is connected to the hotsource of the reactor. HEiDi usesthe higher flux of neutrons withshorter wavelengths (in compar-

ison to thermal neutrons) to ob-serve a larger area of the recip-rocal space Q (|Q|= sinΘmax/λ ).The gain in observable Bragg re-


flections yields a significant en-hancement in detailed structuralinvestigations. Examples for thisimportant application for scien-tific cases after the launch ofFRM-II are structural investiga-tions on high Tc superconduc-tor [1], structural phase transi-tions with atomic or molecular or-der/disorder character [2], anhar-monic behaviour and the impor-tant role of hydrogen bonds, e.g.in ferroelectrics of the KDP fam-ily. Due to the bandwith fromλ = 0.3Åto λ = 1.1Åit is alsopossible to investigate the quan-titative description of the wave-length dependent extinction ef-fect, e.g. for TOF Laue tech-niques at spallation sources [3].An overview of the instrument isshown in picture2.3.

Details of the instrument andits applications were presentedon the german conference forcrystallographers, DGK 2003, inBerlin and at the ECNS confer-ence in Montpellier.

In 2003 the following compo-nents of the instrument were com-pleted: The three primary beamcollimators (manufactured at theHMI in Berlin) were mountedin the beam shutter of SR-9.They define the horizontal di-vergence of the primary beam

Figure 2.3:Overview

Figure 2.4:Instrument in experimental hall

(15’, 30’, 60’). They are neededfor the choice of the best com-promise between good resolutionand high intensity.

The position of the biologi-cal shielding was tested and re-aligned to the direction of beamtube SR-9B. Two beam shutterwere integrated in front and be-hind the monochromator pit. Inpicture 2.4 the biological shield-ing with open monochromator pitand the diffractometer unit itselfwith open detector shielding (tomount the single detector)

are shown.One Cu-(220) and one Cu-

(420) crystal were cut into slicesat the FZ Jülich to mount themon the focussing units of themonochromator. Recently thealignment of the slices was suc-cessfully tested at the single crys-tal diffractometer SV28 (DIDO,FZ Jülich). A picture of one fo-cussing unit with a mounted Cucrystal is shown in picture2.5.

Another focussing unit for theGe-(311) crystal was delivered atthe end of 2002. The thermoplas-


tic crystal treatment in Risoe hasbeen found to be the best way toget suitable Ge monochromatorcrystals. This third monochroma-tor unit should be ready at the endof 2004.

Concerning the optional 2D-detector a small (200*200 mm2

sensitive area)3He wire chamberwith high pressure (about 8 bar)was found to be the most suit-able technique for our needs. Thecomplete system should be readyin spring of 2004.

In summary all major compo-nents of the instrument are nowavailable. After the final tests andfine tuning of the instrument dur-ing the first reactor period we ex-pect this instrument to be readyfor many scientific cases in struc-tural science. Figure 2.5:Focussing unit with mounted Cu-(220) crystal

[1] Braden M., Meven M., Reichardt W., Pintschovius L., Fernandez-Diaz M., Heger G., Nakamura F.,Fujita T. Phys. Rev. B, 63, (2001), 140510.

[2] Schiebel P., Hoser A., Prandl W., Heger G., Schweiss P.J. Phys. I France, 3, (1993), 987.[3] Jauch W., Schultz A., Heger G. (1987).

2.3 MatSci-R – Materials-Science Reflectometer

J. Major 1, J. Franke1, A. Rühm 1, U. Wildgruber 1, H. Dosch1

1 Max-Planck-Institut für Metallforschung Heisenbergstr. 3, D-70569 Stuttgart, Germany

Within the framework of aninitiative of the Max PlanckSociety for the Advancementof Science (MPG) ‘MaterialsScience at the New NeutronSource FRM-II’, this instrumentis built and will be operatedby the Max-Planck-Institut fürMetallforschung (Stuttgart). Anoverview of the instrument hasbeen presented in the 2002 an-nual report of the FRM-II.

By now the design is com-pleted and most hardware com-ponents are built and readyfor installation at the cold neu-tron guide NL-1 in the neutron-guide hall of the FRM-II. Themonochromator shielding (Fig.2.6) is in place and the neu-tron guide is operational. Themonochromator (Fig. 2.7) isready to be tested in the neutronbeam.

The length of the monochro-mator is determined by thewidth of the neutron guide(60mm) and the monochroma-tor Bragg angle for 2 Å neu-trons (17.34◦). During com-missioning of the instrumentup to eleven HOPG (highly ori-ented pyrolytic graphite) plates(approx. 66 mm× 21.5 mm ×2 mm) will be used. Each crystalis directly mounted to a geared


Figure 2.6:Monochromator shielding at the cold neutron guide NL-1 in theFRM-II guide hall with sliding front wall (including the rotating drum)in place.

down miniature ARSAPE step-ping motor (Faulhaber, Schö-naich, Germany) with an angu-lar step size of 0.0005◦. Laterthe single plates can be replacedwith stacks of three HOPGplates introducing a small an-gular offset between three theindividuals to achieve a bettermatch between the angular ac-ceptance of monochromator andprimary beam divergence andthus increase the intensity ofthe monochromatic beam. TheHOPG monochromator “crys-tals” are lowered from above inthe “white” beam of NL-1. Nl-1has a cross section of 120 mm×60 mm (h×w). Due to the lowbeam-center line, we are plan-ning to use the upper half of thebeam to extract monochromaticneutrons.

Furthermore, the monochro-mator device will rotate with re-spect to the incoming beam whenthe sample table is moving up or

down. This assures maximummonochromatic intensity at thecenter of a ‘liquid mode’ sampleeven at large angles of incidence.The motor controllers (MPI de-sign) to move the monochro-mator crystals, the slit systems,etc. are compact rack mount-able units (19′′ × 2 rack units)which can be equipped with up tothree ‘Sixpack’, ‘Quadpack’, or‘Monopack’ modules (Trinamic,Hamburg, Germany). The field-bus interface is an RS485 sys-tem (with optional CAN proto-col). Most of the other con-trol hardware (larger stepper mo-tor controllers, angular and lin-ear encoders, temperatures sen-sors, etc.) is integrated via the‘Profibus-DP’ protocol and addi-tional gateways to other fieldbussystems if necessary.

Currently all ‘crystals’ arecharacterized by means ofgamma-ray diffraction (MPI,Stuttgart). This is necessary be-

cause the overall mosaicity of theHOPG plates varies significantly.Furthermore, depending on thegeometrical cross section of themonochromatic beam, differentparts of the ‘crystals’ will beused. This requires to measurethe dependency of the mosaicspread along the ‘plates’ in or-der to be able to optimize themonochromator device numeri-cally. The variation of the mosaicspread for the eleven center crys-tals is between 0.37◦ (fwhm) and0.71◦ (fwhm). Although this vari-ation is within the specifications(0.8◦ ± 0.2◦) of the purchasedHOPG-ZYB (Advanced Ceram-ics, a GE company, USA), it is solarge that additional Monte CarloRay tracing calculations are nec-essary to optimize the arrange-ment of all HOPG plates withinthe monochromator. This is espe-cially important for shorter wave-lengths where the spatial neutronflux distribution at the exit of NL-1 is particularly complex due tothe curved design of the guidewith different ‘m-value’ for bothside walls.

Figure 2.7: The MatSci-Rmonochromator gallows.


Detector electronics (GKSS,Geesthacht, Germany) and activeanti-vibration hardware (Halcy-

onics, Göttingen, Germany) willbe put into operation in January2004. The final assembly of the

reflectometer will start in Febru-ary 2004.

2.4 REFSANS – Neutron Optics of the Reflectometer REFSANSfor Comprehensive Investigations on the Air/Liquid Interface

R. Kampmann1, M. Haese-Seiller1, V. Kudryashov 1,3, Ch. Daniel2, V. Deriglazov3, B. Toperverg3,A. Schreyer1, E. Sackmann2

1 GKSS Forschungszentrum, Institut für Werkstoffforschung, D-21502 Geesthacht, Germany2 Physik Department E22, TU München3 Petersburg Nuclear Physics Institute, Gatchina, 188350, Russian Federation

Introduction

REFSANS is a time-of-flight(TOF) instrument with three dou-ble disc choppers which allowsetting the wavelength resolutionin the broad range of 0.25%<∆λ/λ < 15 % to meet the differ-ent experimental demands com-prising high and low resolu-tion reflectometry and small-angle scattering at grazing inci-dence (GI-SANS) [1–3]. The in-strument is being built at the endof the cold neutron guide NG-2b with 12 mm high and 170 mmwide beam cross-section. Thisunusually broad beam is used to

a)

sm: m~2hNG-2b

w

Figure 2.8: Cross-sections of neu-tron guide elements (NGE) ofREFSANS, perpendicular to thebeam direction (a: NGE-3 and -4;b: NGE-7 and -8).

install radially collimating neu-tron optics to substantially im-prove the GI-SANS capability ofREFSANS.

Neutron optical design

REFSANS is equipped withnovel Neutron guide elements(NGE) which allow continuingthe beam of NG-2b throughREFSANS (upper channels inFig. 2.8a and b), collimating thebeam vertically while maintain-ing the horizontal divergence(lower channel in Fig. 2.8b),

Figure 2.9:Schematic view of REFSANS and its neutron optics, details arementioned in the text.

reflecting the beam from top tobottom (or vice versa) onto thesample surface by means of theupper channels of NGE’s-3, -4,-7, and -8 (Fig.2.8 and upperchart in Fig.2.9). A further pos-sibility for declining the beamat an angle of≈ 180 mrad isprovided by a special bender(see beam B4 in upper chart ofFig. 2.9). Finally, the lower chan-nels of NGE-3 and -4 (Fig.2.8a)may be moved into the beamto separate and pre-collimate13 partial beams which are fo-cused in the detector plane at a


Figure 2.10:Neutron optical bodyin the beam guide chamber withthe mechanics for vertically align-ing a NGE and an ancillary di-aphragm.

distance of≈ 9 m from the sam-ple. The separation of the latteris maintained in the beam guidechamber over≈ 8 m by ancillarydiaphragms with comb-like aper-tures (Fig.2.9, central chart andFig. 2.10).

The intensity of such radiallyfocused beams hitting a samplewith a size of≈ 60 mm×60 mmhas been calculated on the ba-sis of the brilliance of the coldsource of FRM-II and accountingfor the expected transmission ofthe chopper, the neutron opticsand the radially collimating sys-tem [3]. An excellent peak inten-sity is calculated for the case of aresolution in momentum transferof ∆q/q≈ 14% (Fig.2.11).

It is pointed out that this re-sults on the one hand from theexcellent transmission of the 3-double-disc chopper which isclose to that of a 10 % selectorand on the other hand from fo-

cussing of 13 beams in the de-tector plane. This altogether re-sults in an intensity which wouldbe expected if 13 modern SANSfacilities at a high flux reactorwould be arranged in focussingand grazing incidence geometry.

Due to this very high beam in-tensity at grazing incidence REF-SANS is expected to offer newGI-SANS perspectives. Today’smeasuring times could be de-creased by an order of magnitudeand time dependent GI-SANSstudies will become feasible. Allinvestigations may be performedat the air-water interface withhorizontally aligned samples andmay be combined with furthermeasurements of the specular re-flectivity to characterise compre-hensively vertical and horizontalsurface structures.

The intensity may be fur-ther increased by almost oneorder of magnitude by increas-ing the vertical beam divergence

Figure 2.11:Calculated peak and time averaged beam intensitiesIpeak andImean of 13 beams horizontally focused in the detector plane (details inthe text).

(Fig. 2.11). This can easilyachieved by guiding the 13 par-tial beams in the upper channelsof the neutron guide elements inthe beam guide chamber (Fig.2.8and2.9). This geometry will es-pecially allow of measuring ex-tremely weak diffuse surface scat-tering from very small lateralstructures and of following thetails of diffuse surface scatteringpatterns.

Acknowledgements

The great contribution of thetechnical department of GKSSto constructing and manufactur-ing of REFSANS components isgratefully acknowledged. Thedevelopment of REFSANS hasbeen supported by the Ger-man Federal Ministry of Edu-cation, Research, and Technol-ogy (BMBF) under Contracts 03-KA5FRM-1 and 03-KAE8X-3.

[1] Kampmann R., Haese-Seiller M., Marmotti M., Burmester J., Deriglazov V., Syromiatnikov V.,


Okorokov A., Frisius F., Tristl M., Sackmann E.Applied Physics A, 74, (2002), 249–251.[2] Kampmann R., Haese-Seiller M., Kudryashov V., Deriglazov V., Tristl M., Daniel C., Toperverg B.,

Schreyer A., Sackmann E.Applied Physics A. In press.[3] Kampmann R., Haese-Seiller M., Kudryashov V., Deriglazov V., Tristl M., Schreyer A., Sackmann E.

Nuclear Instruments and Methods A. Submitted.

2.5 2D-multi-wire neutron detector with a large sensitive areaand high spatial resolution for SANS and reflectometry

R. Kampmann1,2, M. Marmotti 2, M. Haese-Seiller1, V. Kudryashov 1,3

1 GKSS Forschungszentrum, Institut für Werkstoffforschung, D-21502 Geesthacht, Germany2 DENEX Detektoren für Neutronen und Röntgenstrahlung GmbH, D-21339 Lüneburg, Germany3 Petersburg Nuclear Physics Institute, Gatchina, 188350, Russian Federation

Introduction

A two-dimensional position-sensitive multi-wire gaseous de-tector for small-angle neutronscattering (SANS), reflectome-try and high-resolution diffrac-tometry has been developed atthe GKSS research centre in co-operation with DENEX GmbH.The counter with a sensitive areaof 500× 500 mm2 has been de-signed to be used in the reflec-tometer REFSANS being built atthe new high flux reactor FRM-II[1].

Figure 2.12:REFSANS detector be-ing tested at SANS-2 at GKSS

Design and performanceof the detector

The REFSANS detector is a3Heand CF4 filled multi-wire propor-tional counter (MWPC) with de-lay line read-out, its design isbased on that of detectors withsmaller active areas developedpreviously at GKSS [2, 3]. REF-SANS needs a detector whichmeets requirements both of re-flectometers and SANS instru-ments [1]:

• Sensitive area:500× 500 mm2;

• Spatial resolution:≈ 2 mm× 2 mm;

• Wavelength range:0.3 nm< λ < 3 nm;

• Read-out:fast, ToF-application;

• Background:very low;

• γ-sensitivityεγ :extremely low (εγ �10−6).

All electrodes are made oftungsten wires with a distanceof 2 mm. Drift electrodes limit

the detection volume which has adepth of 3 cm to achieve a highdetection probabilityp at ratherlow 3He partial pressure. The an-ode is located in the centre of thedetection volume. Its distanceto the pick-up electrodes is only5 mm to achieve high capabilitiesfor both good position resolutionand high local count rates.

One prototype of REFSANSdetector was manufactured andcharacterised with neutrons. Ithas been filled with only 0.8 bar3He and 1 bar CF4 for use in theSANS-2 at the GKSS ResearchCentre (Fig.2.12).

The detection probability wasmeasured with a well collimatedneutron beam at SANS-2 in com-parison with an almost black3He counter tube. A detectionprobability larger than 50 % overthe range for 0.5 nm< λ <1.8 nm with a maximum of≈58 % at λ = 1.0 nm was found(Fig. 2.13). This detection prob-ability is close to the ideal onewhich is calculated by account-ing for the losses in the detectorwindow (20 mm thick Al-alloy)


and the dead volume between thewindow and the first drift elec-trode (thickness:≈ 2 mm).

The γ-efficiency was mea-sured without changing the set-ting of the analogue electron-ics. Radioactive sources werepositioned at short distance (≈10 cm) in front of the detec-tor. Extremely lowγ-sensitivitiesof εγ � 10−7 were measured(Fig. 2.13).

After these tests the detec-tor was mounted in the evacu-ated scattering tube of SANS-2/ GKSS. An isotropic scattererwas set in the beam at a dis-tance of≈ 3 m from the detec-tor which was covered by a Cd-mask (≈ 20 cm in front of its win-dow) with a grid of holes with dis-tances of 50 mm. A spatial res-olution of ≈ 3 mm × 3 mm to-gether with an excellent linearitywas found [3].

Summary

Due to its large sensitive area(500× 500 mm2), its spatial res-olution of ≈ 3 mm× 3 mm andits low γ-sensitivity (εγ � 10−7)

the prototype detector meets al-ready central requirements of theREFSANS detector and it is anexcellent detector for a SANSat a monochromatic beam. Forthe use in REFSANS the detec-tion probability will be furtherincreased by changing the elec-trode design to avoid any deadvolume behind the detector win-dow. Furthermore, the spatialresolution will be improved to≈ 2 mm × 2 mm by increas-ing the partial pressure of CF4.The REFSANS detector has beenlicensed by GKSS to DENEX.

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1,0

0 5 10 15 20 25 30 35 40 45

wavelength [Å]

prob

abili

ty

absorption of detection volumecalc. ideal detection probabilityas measured

Gamma efficiency1. εγ = 3 × 10–8 with γ-60Co, 1.32 MeV

2. εγ = 2 × 10–8 with γ-137Cs, 622 keV

Figure 2.13:Neutron detection probability andγ-efficiency of the prototypeREFSANS detector.

Acknowledgments

The support of M. Pauls, H. Eck-erlebe and K. P. Pranzas dur-ing tests of the REFSANS de-tector at SANS-2 / GKSS isgratefully acknowledged. Thedevelopment of REFSANS hasbeen supported by the Ger-man Federal Ministry of Edu-cation, Research, and Technol-ogy (BMBF) under Contracts 03-KA5FRM-1 and 03-KAE8X-3.

[1] Kampmann R., Haese-Seiller M., Marmotti M., Burmester J., Deriglazov V., Syromiatnikov V.,Okorokov A., Frisius F., Tristl M., Sackmann E.Applied Physics A, 74, (2002), 249–251.

[2] Marmotti M., Burmister J., Haese-Seiller M., Kampmann R.Nuclear Instruments and Methods A,477, (2002), 347–352.

[3] Kampmann R., Haese-Seiller M., Marmotti M., Burmester J., Deriglazov V., Syromiatnikov V.,Okorokov A., Frisius F., Tristl M., Sackmann E.Applied Physics A, 74, (2002), 252–254.


2.6 RESI – The Single Crystal Diffractometer

B. Pedersen1, G. Seidl1, W. Scherer2, F. Frey3

1 ZWE FRM-II, TU München2 Inst. f. Physik, Universität Augsburg3 Sektion Kristallographie, Geodepartment, LMU, München

Introduction

The thermal single crystal diffrac-tometer RESI has been designedto record strong and weak scatter-ing in large portions of Q-spacewith high resolution. HenceRESI is optimized for both (i)chemical problems requiring fastscanning of the reciprocal spaceand (ii) crystallographical de-mands, where non-Bragg scatter-ing is of major importance.

This can be achieved by a flex-ible setup with different inter-changeable monochromators, go-niometers and detector systems.

Status

RESI is now in the final test-ing phase. All essential compo-nents, which are needed to oper-ate RESI in the standard modus,have been installed and testedduring the last year.

Also new developments, suchas the construction of a 511-Ge monochromator could be fin-ished in 2003. The 511-Gemonochromator crystals, consist-ing of stacks of soldered andbent wafers, were processed byRisø and delivered in the mean-time. The effective mosaic hasbeen optimized in the same wayas for the copper crystals byMonteCarlo-simulations to getan optimum flux/resolution ratio

[1]. The final characterisation isplanned for spring time next yearupon availability of neutrons.

The goniometer software haspassed all tests, and is now readyfor operation. It is based onthe Collect™ package by Bruker-Nonius and allows easy deter-mination of cell parameters andorientation matrices, calculationof optimised scans to cover thereciprocal space and is comple-mented by an advanced and user-friendly integration software forBragg peaks.

The second goniometer is de-signed and in the final assemblystage (see Fig.2.14). It allows

Figure 2.14:The goniometer for heavy sample environment

the usage of heavier sampleenvironments (e.g. furnances,cryostats) in combination withthe image plate detection system.

Outlook

During the next year we will in-stall the heavy goniometer andstart to operate RESI to performthe “hot” test of the instrumentand first measurements to charac-terise the instrument.

The analyzer option, whichallows the separation of inelas-tic and elastic contributions, isscheduled for installation for fall2004.

[1] Pedersen B., Seidl G., Scherer W., Frey F. .FRM-II Annual Report, 16–19.


2.7 SPODI – Structure Powder Diffractometer

R. Gilles1, M. Hoelzel2, M. Schlapp2, B. Krimmer 2, H. Boysen3, H. Fuess2

1 ZWE FRM-II, TU München2 Fachbereich Materialwissenschaft, Technische Universität Darmstadt3 Ludwig-Maximilians-Universität München

The design and status ofthe new Structure POwderDI ffractometer (SPODI), cur-rently being built up at the newneutron source FRM-II in Garch-ing near München is reviewed.Based on an optimization of thecomponents by Monte Carlo sim-ulations, new features includesupermirror neutron guides, avery high monochromator take-off angle and an array of posi-tion sensitive detectors. Thisconcept aims at an optimum com-promise of improved resolution,higher intensity, better profileshape and lower background.The design of new instrumentscan benefit from recent theo-retical and technical progress

Figure 2.15:Side-view on instrument SPODI with shielding of the neutronguide & monochromator and detector in the foreground

as well as new ideas. In orderto be able to solve complicatedstructures reliably by powderdiffraction, not only a high res-olution and intensity are essen-tial. One also requires goodpeak profiles that can be de-scribed as perfectly as possible.For strongly overlapping reflec-tions, uncertainties in the exactshape can lead to a wrong distri-bution of the intensities. More-over, such knowledge virtuallyincreases the resolution of theinstrument far beyond that givenby the widths of the peaks. Sim-ilar arguments hold for a goodpeak to background ratio whichcan be achieved not only by alow background, but also by

narrow tails of the peaks. Finally,such good peak shapes should bemaintained up to large scatteringangles 2θ . Based on these con-siderations computer simulationshave been used to optimize eachsingle component and addition-ally their interaction along the in-strument. The resulting concepttogether with test measurementsof single components at neutronsources have been described inreferences [1–5]. The instrumentSPODI is shown in Fig.2.15.

During his visit at the FRM-II on the occasion of the grant-ing of the third partial license, theBavarian Minister-President Dr.Edmund Stoiber participated in apresentation about the StructurePowder Diffractometer SPODI(Fig. 2.16). Under developmentis the construction of a small-angle scattering apparatus in for-ward direction to characterize thestructure and the size distributionof powder samples. Monte Carlosimulations have been carried outto optimize this additional fea-ture [6]. Other applications suchas the determination of particlemorphology, texture analysis andinvestigations on the crystallinityof samples are planned with theimage plate system of this SANSoption.


Figure 2.16:The Bavarian Minister-President Dr. Edmund Stoiber visits theinstrument SPODI

[1] Gilles R., Artus G., Saroun J., Boysen H., Fuess H.Physica B, 276-278, (2000), 87.[2] Gilles R., Artus G., Saroun J., Boysen H., Fuess H. In:Proc. HERCULES X EuroConference

(Grenoble, 2000).[3] Gilles R., Krimmer B., Saroun J., Boysen H., Fuess H.Mat. Sci. For., 378-381, (2001), 282.[4] Gilles R., Krimmer B., Boysen H., Fuess H.Appl. Phys. A 74 (Suppl.), 74, (2002), 148.[5] Hoelzel M., Gilles R., Schlapp M., Boysen H., Fuess H. (2003). Submitted toPhysica B.[6] Schlapp M., Hoelzel M., Gilles R., Ioffe A., Brueckel T., Fuess H., von Seggern H. (2003).

Submitted toPhysica B.

2.8 STRESS-SPEC – Materials Science Diffractometer

M. Hofmann 1, G. Seidl1, R. Schneider2

1 ZWE FRM-II, TU München2 Hahn-Meitner-Institut, Berlin

Introduction

The Materials Science diffrac-tometer Stress-Spec is situatedat the thermal beam tube SR3and will be primarily used forresidual stress analysis and tex-ture measurements of new materi-als and engineering components.Its main characteristics have been

described elsewhere [1, 2] andhere only the developments oflast year will be reviewed.

Test phase at HMI Berlin

In cooperation with the Hahn-Meitner-Insitut, Berlin the sec-ondary part (monochromator lift,sample table, detector and elec-

tronics) of the materials sciencediffractometer STRESS-SPEChas been moved to the BER-IIreactor and installed at beamport D1S/E7 during the last year.This allowed to test and optimizethose components, to gain expe-rience with new calibration andalignment procedures and to runfirst experiments.


For instance in one of thefirst measurements the relax-ation of residual stresses in coldrolled crank shafts under dif-ferent mechanical load condi-tions was investigated in coop-eration with Volkswagen AG (P.Lemke, HMI, Diploma thesis2003). Other examples includedexperiments to study magneto-viscosity effects in some ferroflu-ids materials like nanocrystallineCoFe2O4 (E. Steichle, TUM),temperature dependent residualstress analysis in SiC reinforcedAlSi metal composites (U. Gö-bel, Siemens AG) and resid-ual stress analysis in as castAl bars (T. Hanss, UTG-TUM).Also investigated was a proto-type rocket nozzle fabricated byAstrium GmbH (figure 1) us-ing a carbon fibre reinforced ce-ramic composite material (J. Re-belo Kornmeier, TUM). Here the

Figure 2.17:Example of a typical measurement setup at Stress-Spec.

main interest was focussed onwhether and how residual stressdue to different thermal expan-sion of the materials when heatedcan be relieved by micro cracksin the matrix material. This isin turn could help to counteractfailure or fatigue of the compo-nent and extend its life time un-der working conditions.

In course of this test phasesome parts of the equipment weremodified due to the experiencegained using the diffractometerin real life conditions. As anexample a new detector slit sys-tem has been installed with re-gard to reproducibility of the slitpositioning and sturdiness. Alsoa new high capacity Eulerian cra-dle (sample load up to 50 kg) hasbeen designed at the HMI, whichwill be available in early sum-mer 2004 for routine operation atFRM-II. In addition most of the

electronics and mechanical sys-tems could be tested and the dataacquisition software has beenadapted and upgraded.

FRM-II and outlook

While the diffractometer was inBerlin work continued on theshielding at FRM-II, with devel-opment and installation of newfast shutter systems, new primarybeam slits and mobile block sys-tem electronics. With deliveryof the focussing device for theGe(511)-monochromator earlyWinter 2003, two monochroma-tors will now be available atstartup of the reactor FRM-II.As a third monochromator op-tion we will install a PG(002)-monochromator, which will bemainly used for texture analy-sis and high intensity measure-ments. This monochromator willbe available in spring 2004. Itis forseen to transfer the diffrac-tometer back from HMI Berlinto the beam port SR3 at theFRM-II reactor in January 2004.

[1] Mayer H. M., Pyzalla A., Reimers W.Mat. Sci. Forum, 29, (2000), 347–349.[2] Hofmann M., Seidl G. A., Schneider R.Ann. Report FRM-II.

38 3 Spectrometers

3 Spectrometers

3.1 MIRA – Very Cold Neutrons for New Methods

R. Georgii1, N. Arend 2, P. Böni2, H. Fußstetter1, D. Lamago1, G. Langenstück1, H. Wagenson-ner 1

1 ZWE FRM-II, TU München2 Physik-Department E21, TU München

The instrument MIRA (seeJB 2002) in its current config-uration provides neutrons inthe wavelength range 8 to 15A using a mica monochroma-tor. The instrument has beendesigned for being able to mea-sure a great variety of phenom-ena with very cold neutrons.

Figure 3.1:The reflectometer option of MIRA without the detector and themagnets for the guide field.

The reflectometer for meso-scopic structures is equippedwith a polarised option. Neu-trons will be polarised usinga switchable transmission mul-tilayer polarizer. Then a fieldof approximately 50 Gauss pro-vided by permanent magnets willguide the polarised beam to the

sample. Here the standard FRMsample environment (cryostats,cryomagnets and ovens) can beused. At last the neutrons areagain guided by a magnetic fieldto the detector. Currently a3Hedetector is being used. A 2-dimenisenional position sensitivedetector with 20 x 20 cm sensi-tive area and a resolution of about1 mm has been bought fromGKSS and is currently tested athe reactor in Geesthacht. It willbe available early next year.

The instrument setup for thisoption has been successfully in-stalled (see Fig. 3.1). It isfully operational, with the excep-tion of the analysing possibility,which will be added early nextyear. The instrument control soft-ware exits, but is still on a very ba-sic level. Nevertheless first reflec-tivity measurements could be per-formed. A simple evaluation soft-ware has been developed and willbe available after testing with realdata also early next year.

3 Spectrometers 39

3.2 Phonon Lifetime Measurement Using the NRSE-TASTechnique

T. Keller 1, K. Buchner 1, R. Golub2, K. Habicht 2, H. Klann 1, F. Mezei2, M. Ohl 1, B. Keimer 1

1 Max-Planck-Institute for Solid State Research, Stuttgart2 Hahn-Meitner-Institut, Berlin

Introduction

The NRSE-TAS spectrometer atthe FRM-II is a combination ofthe resonance spin echo(NRSE)and the triple axis (TAS) spec-troscopy techniques. This instru-ment will allow the determina-tion of the energies and lifetimesof dispersive excitations, includ-ing both phonons and magnons,over the entire Brillouin zone.The design [1] of the NRSE-TASat the FRM-II (Fig. 3.2) isbased on a prototype instrumentset up at BENSC, Berlin [2]. Atthis instrument we first demon-strated the spin-echo phonon-focusing technique [3]. As anexample for this technique we

Figure 3.2:The NRSE-TAS spectrometer at the FRM-II

show here the determination oflifetimes of transverse acousticphonon modes in Pb at temper-atures between 5-300 K at lowq values. To correct for instru-mental resolution effects, an an-alytical resolution function forthe NRSE-TAS technique [4] hasbeen developed.

The polarizationP measuredby a combination of a spin echoand a TAS instrument is given by

(3.1) P ∝∫

S(Q,ω)T(k i ,k f )exp[iφ(ki ,kf )]d3kid3kf

where S(Q,ω) is the scatteringfunction, T(k i ,k f ) is the trans-mission probability of the TAS

spectrometer.φ = φi − φ f is thetotal Larmor precession angle forneutrons passing precession fieldregions before and after the sam-ple.

Mezei showed [5] that in orderto measure phonon linewidths,the spin echo instrument has tobe tuned both to the phononenergy ω0 and group velocity∇qω0(q0). With our resonancespin echo technique, this tuning

is achieved by rotating the RFspinflippers bounding the preces-sion regions thus that the normalvectors on the coil surfaces sat-isfy the condition(3.2)

ni, f ‖ [(h̄/m)kI ,F −∇qω0(q0)]

The RFνi, f applied to the coils inthe first and second arm is tunedto the ratio(3.3)

νi

ν f=

k2I

k2F

h̄kI −m∇qω0(q0)∣∣h̄kF −m∇qω0(q0)∣∣

kI ,F are the mean wavevec-tors. For non-dispersive excita-tions (∇qω0(q0) = 0) the tuningconditions (eqs.3.2, 3.3) requireto set the field boundaries per-pendicular to the neutron beamand to set the frequency ratio toνi/ν f = (ki/kf )3.

If we assume the scattering

40 3 Spectrometers

law S(Q,ω) to be independent ofQ in that part of the dispersionsurface which is located withinthe TAS resolution ellipsoid, wecan integrate over the momentumcomponents in eq. (1) and obtain

(3.4) P(τ) ∝∫

S(∆ω)T(∆ω)exp(−iτ∆ω + iφ0)d(∆ω)+c.c.

In a high resolution spin-echoexperiment the energy resolutionT(∆ω) is usually much broaderthan S(∆ω),thus T(∆ω) can beconsidered to be constant overthe integration range. Then eq.3.4 directly yields the cosineFourier transform of the line. Fora Lorentzianlineshape with half-width-at-half-maximumΓ the po-larization follows a simple expo-nential decay

(3.5) P(τ) ∝ exp(−Γ · τ)

τ is thespin echo time

(3.6) τ =A1,2

kI ,F ·ni, f [(h̄/m)k2I ,F −kI ,F∇qω0(q0)]

whereAi, f = (2πm/h̄)νi, f Li, f cosθi, f ,Li, f are the lengths of the pre-cession regions andθi, f are theinclination angles of the RF coils.Note thatτ is identical to the cor-relation timet which is the funda-mental variable in the time depen-dent van Hove density-densitycorrelation function.

If the energy width ofS(∆ω)is comparable to or broader thanT(∆ω), eq. 3.5 gets much morecomplicated. In this case itseasier to determine the linewidthwith a standard TAS scan.

Experiment

Inelastic neutron scattering ex-periments were performed at thecold triple-axis spectrometer V2(FLEX) at the BER-II research

reactor, HMI, Berlin. This instru-ment can optionally be equippedwith an NRSE setup [3]. Thespectrometer was configured intheSM=−1, SS=−1, SA= +1scattering senses with fixedkI =1.7Å−1. The PG monochromatorwas vertical focussing, the ana-lyzer (PG) was flat. A tunable PGfilter [6] was installed to suppresssecond order contamination.

In order to satisfy the spin-echo conditions (eqs.3.2, 3.3)the slope of the dispersion has tobe known. This quantity has beenobtained from a Born-von- Kàr-màn fit to published dispersiondata [7, 8]. Typical coil tilt angleswereθ1,2 ' 30, typical frequencyratiosν1/ν2 ' 1.2 .

For each phonon mode andtemperature the TAS was set tothe nominal(Q,ω) values whereit was kept fixed during spin-echoscans. The polarisation was mea-

sured for at least four differentvalues of the spin echo timeτper phonon and temperature (Fig.3.3 ). Within statistical error ourmeasurements are in agreementwith a single exponential decay(eq. 3.5). Thus we find noevidence for possible deviationsfrom Lorentzian lineshapes.

To correct for resolution ef-fects, we used an analytical res-olution function for the NRSE-TAS instrument [4] taking intoaccount instrumental parametersand sample properties (mosaic-ity and curvature of the disper-sion surface). Like in quasielas-tic spin echo experiments [9], themeasured polarizationPexp(τ)can be written as a product ofa resolution termPres(τ) and theFourier transform of the scatter-ing law:

(3.7) Pexp(τ) = Pres(τ)×∫

S(∆ω)exp(−iτ∆ω)d(∆ω)

The resolution calculation wastested for a phonon at very lowq and low temperature, which isexpected to have zero linewidthon our resolution scale. Thusthe polarisation should be con-stant (P(τ) = 1). The experi-mental data in Fig.3.4 shows adecaying polarisation, which re-sults both from sample mosaic-ity ηS and curvature of the dis-persion surface. For increasingq, the curvature of the disper-sion surface gets smaller, the in-fluence of the mosaicity staysnearly constant. For longitudi-nal phonons (not studied in thepresent experiment), the depolar-izing effect of ηS enters only insecond order and is negligible

3 Spectrometers 41

10 20 30 40 50

0,1

1

Pol

ariz

atio

n

τ [ps]

Figure 3.3: Polarization vs. spinecho timeτ for the [2 0.1 0] TAphonon in Pb, T = 200 K. The lineis a fit to an exponential decay.

on our resolution scale, but varia-tions of the lattice spacing∆d/dbecome a first order effect andhave to be considered.

In 3.6 the corrected linewidthsof our measurements are com-pared to the TOF experiment ofFurrer and Hälg [8] to demon-strate the enhanced resolution ofthe NRSE-TAS technique.

Since lattice defects are not ex-pected to contribute to the mea-sured linewidths only electron-phonon and phonon-phonon in-teraction remain. Ab initio cal-culations provided by the MPIStuttgart [10] show that theelectron-phonon contribution to

0 20 40 60

0,1

1

Pol

ariz

atio

n

τ [ps]

Figure 3.4: [0.025 0.025 0] TAphonon in Pb, T = 50 K. Cal-culated depolarization resultingfrom sample mosaicity (dotted)ηS = 4.2′ (FWHM), curvature ofthe dispersion surface (dashed)and the combined effect of both(solid). Open circles: experimen-tal data of a linewidth measure-ment performed at BENSC.

0 10 20 30 40

0,1

1

Pol

ariz

atio

n

τ [ps]

0 10 20 30 40 50

τ [ps]

Figure 3.5: Calculated depolar-ization due to sample mosaic(dashed), curvature of the dis-persion surface (dotted) and thecombination of both (black line).Experimental raw data (circles)and normalized data (squares).Left: [2 0.05 0] TA phonon atT = 300 K. Right: [2 0.1 0] TAphonon at T = 300 K.

the phonon linewidths is negligi-ble. Furthermore we have mea-sured the linewidth of the [0.10 0] TA mode at temperaturesabove and belowTC. No changein linewidth was experimentallyobserved.

Zoli [11] has calculatedlinewidths for Pb based on apseudopotential model and ob-tains much larger values thanobserved in our measurement(Fig. 3.6). He obtained theharmonic constants from a fitto experimental phonon disper-sion data and macroscopic elastic

0,0 0,1 0,2 0,30

20

40

60

Γ [µ

eV]

ξ [r.l.u.]

Figure 3.6:Phonon linewidths of the[ξ00] TA mode in Pb. Solidcircles : NRSE experiment (thiswork) ; red open circles : TOF ex-periment, Furrer and Hälg, 1970

constants. Anharmonic constantswere obtained from a fit to thelinear coefficient of thermal ex-pansion and the thermodynamicGrüneisen-parameter. There is ofcourse general doubt over the ac-curacy of the anharmonic param-eters obtained from a pseudopo-tential model. The fact that theinteratomic potential in the fccmetals is of long range implies alarge number of parameters. Fur-thermore an approach with onlytwo-body forces is not possiblesince the macroscopic elastic con-stants do not satisfy the second-order Cauchy relations. In thelast years however there is agrowing number of publicationswith ab initio models of the inter-atomic interaction which can pro-vide the Fourier transforms of thehigher order derivatives of the in-teratomic potential. No such cal-culations are published for Pb.

Conclusions

Using the NRSE technique on acold TAS we have sucessfullyapplied the phonon-focusingtechnique to dispersive phononmodes in single crystallinelead. The broadening of theone-phonon peaks resulting inLorentzian linewidths could bemeasured with high accuracy su-perior to conventional neutronscattering techniques used inprevious experiments. Phononlinewidths were extracted by tak-ing mosaicity and curvature ofthe dispersion explicitly into ac-count with a resolution functionbased on the Gaussian approx-imation of the TAS transmis-sion function. The result of thelinewidth measurement supports

42 3 Spectrometers

the disagreement between thepseudopotential theory of Pb andprevious experiments.

[1] Keller T., Habicht K., Klann H., Ohl M., H. Schneider B. K.Appl. Phys., A74 [Suppl.], (2002),332.

[2] Keller T., Gähler R., Kunze H., Golub R.Neutron News, 6, (1995), 16.[3] Keller T., Habicht K., R.Golub. In: Mezei F., Pappas C., Gutberlet T. (Editors),Lecture Notes in

Physics(Springer, 2003).[4] Habicht K., Keller T., Golub R.J. Appl. Cryst., 36, (2003), 1307.[5] Mezei F. Inelastic Neutron Scattering(IAEA, Vienna, 1978).[6] Vorderwisch P., Hautecler S., Mezei F., Stuhr U.Physica B, 213-214, (1995), 866.[7] Brockhouse B., et al.Phys. Rev., 128, (1962), 1099.[8] Furrer A., Hälg W.Phys. Stat. Sol., 42, (1970), 821.[9] Mezei F. (Editor).Neutron Spin Echo(Springer, 1980).

[10] Bose S. MPI Stuttgart, private communication.[11] Zoli M. J. Phys. Cond. Mat., 3, (1991), 6249.

3.3 PANDA – The Cold Neutron Three-axes Spectrometer

P. Link 1, D. Etzdorf 1, N. Pyka1,3, M. Rotter 2, R. Schedler2, R. Sprungk2, E. Kaiser2, C. Wetzig2,M. Dörr 2, S. Sahling2, M. Loewenhaupt2

1 ZWE FRM-II, TU München2 Inst. f. Festkörperphysikphysik, TU Dresden3 now at: MOWO Maschinenbau GmbH, Berlin

Personal

30th of April this year Dr.N. Pyka left FRM-II and thePANDA project for a position inindustry. Since 1st of May Dr. Pe-ter Link is the new instrument re-sponsible.

Project status andprogress

During the year assembly of thecomponents of both, secondaryspectrometer as well as the pri-mary spectrometer has been con-tinued. The final positioning ofthe neutron channel in front ofbeam port SR2 was a major mile-stone for the primary spectrom-eter achieved in august. This

included optical adjustment ofall neutron optical devices (col-limators, neutron guide and vari-able horizontal aperture) insidethe channel. All devices havebeen fully integrated in the instru-ments remote control softwareand are ready for operation. Fur-ther all crystal holders for dou-ble focussing Heusler and PG002monochromators and analysershave been manufactured and de-livered. Concerning the sampleenvironment we note the deliveryand successful test of the Oxford15T vertical magnet system in-cluding an adapted Kelvinox3He-4He dilution refrigerator. Thetests after delivery had been per-

formed at PANDA experiment

Figure 3.7:The focussing mechanicsfor the Heusler crystal analyser(foreground) and monochromator(background)

3 Spectrometers 43

Figure 3.8:15T vertical magnet on the PANDA sample table

with the non-magnetic sample-table of PANDA (Fig.3.8). Thesystem reached a maximum fieldof 14.5T and a lowest tempera-ture of about 30mK (as specifi-cation from Oxford Instruments).The cryomagnet is available onPANDA at start of routine oper-ation.

Outlook

During the year 2004 the assem-bly of the main parts of PANDAshould be completed and thespectrometer will be aligned andcharacterised with neutrons be-fore routine operation may startfinally.

3.4 PUMA – The Thermal Three Axes Spectrometer at SR7

K. Hradil 1, H. Schneider1, G. Eckold1, P. Link 2, J. Neuhaus2

1 Institut für Physikalische Chemie, Universität Göttingen2 ZWE FRM-II, TU München

Scientific design

Within a cooperation of the Phys-ical Chemistry Department of theGeorg-August University of Göt-tingen and the Physical Depart-ment of the Technical Universityof Munich the thermal three-axesspectrometer PUMA is installed

at the beam tube SR7. The con-cept of the instrument is an op-timal utilisation of the beam us-ing focusing techniques, reduc-tion of background and higher-order contamination using a ve-locity selector. Thus the in-strument is designed to be welladapted with regard to resolution,

intensity and background for theplanned experiments.

The instrument is particularlyintended for the study of collec-tive excitations like phonons andmagnons. Due to its flexibilityit will also be well suited for theinvestigation of disordered mate-rials, i.e. diffuse and superorder

44 3 Spectrometers

scattering, critical phenomena, orfor time-resolved experiments onnon-equilibrium systems. Fig.3.9 shows the actual setup of theinstrument.

Current status

The detailed technical design ofthe instrument is given athttp://www.frm2.tum.de/puma.This report reviews briefly theprogress of the components in2003. The main componentsfor the single detector option ofthe instrument are in place. Theactivities within this year werefocused on the mechanical align-ment of the spectrometer and thetest of the interplay of differentcomponents. The software con-trol system is being developedand need to be completed in thenear future in combination withthe first tests with neutrons.

Background and high ordercontamination at PUMA will bereduced by the use of a velocity

Figure 3.9:Photograph of the current setup at FRM-II

selector before the monochroma-tor built by Astrium. Due to sometechnical problems the design ofthe selector had to be changedand could not be delivered asscheduled. The container of theselector is rebuilt and the selectordiscs are dynamically balancedby Astrium. The first tests underreal conditions can be performedbeginning of next year and the se-lector can be delivered in the firstquarter of the year.

The horizontal divergence ofthe neutron beam can be definedby remotely controlled Soller col-limators before (20’, 40’, 60’)and after (10’, 20’, 30’, 45’, 60’)the monochromator. Betweensample, analyser and detector thecorresponding collimators haveto be changed by hand.

PUMA is equipped with amonochromator changer for fourdifferent crystals each with amaximum effective area of about460 cm2. PG(002), Cu(220),Cu(111) and Si(311) are

foreseen. While the siliconmonochromator consists of a sin-gle bent perfect wafer, the otherones are made from more than100 small mosaic crystals that aremounted on a doubly focussingdevice. A rocking curve studyof the PG(002) and Cu(220) crys-tals by gamma-rays exhibits amosaicity of around 30’. Thedouble bend apparatus was de-signed within a cooperation ofthe mechanical workshop of theuniversity of Göttingen (IPC) andthe FRM-II and is currently beingtested and aligned in the work-shop of IPC. In the first quar-ter of 2004 the device equippedwith PG(002) crystals will beinstalled at PUMA, tested me-chanically and aligned optically.The installation of the Cu(220)monochromator is forseen after-wards. The Si(311) monochro-mator, built by Swiss Neutron-ics, is aimed for delivery begin-ning of next year. So in fact,PUMA can provide two differentmonochromators during its com-missioning phase. The charac-teristics of both monochromatorswill be tested with neutrons afterstart-up of FRM-II.

Several additional devicessuch as secondary shutter, PGfilter, etc. have been manufac-tured by the workshop of IPC.These components will be deliv-ered end of this year and begin-ning of next year and have to beimplemented afterwards withinthe electronic control system ofthe instrument. Motorized slitsystems with integrated electron-ics as developed in Göttingen areready for use at different posi-tions of the spectrometer. Thesedevices are implemented within

3 Spectrometers 45

the control system of the instru-ment.

PUMA provides two differ-ent options for the sample stage,i.e. a double goniometer (AZ-Systems) for a top load of < 1000kg (in combination with a +/-20mm z-translation table < 150kg load) or an Eulerian cradle, ona 430 mm high base module. Thehigh flexibility concerning the us-able distance to beam (300mm)allows to use complex sample en-vironments as cryomagnets, fur-naces.

Beside the standard FRM-IIsample environment, PUMA pro-vides an closed cycle cryocooler(4K - 300K) which is mountableon the xyz sample stage and op-tionally within the Eulerian cra-dle and a furnace (up to 1200 K)for the Eulerian cradle. Addition-ally, a cryofurnace (3K - 800K)will be delivered by the end of2003. A high temperature fur-nace up to 3000K, which will bebuilt in cooperation with the FZJülich and IPC is planned.

The analyser/detector unitfor the single detector optionof PUMA is completely tested.Moreover, The FRM-II detec-tor group has sucessfully devel-oped a counter module for time-resolved, stroboscopic experi-ments. After the system controlvia TACO is available, the mod-ule can be implemented withinthe instrument control.

The focussing PG(002) anal-yser (Fig.3.10) with a motor con-trolled horizontal and a fixed ver-tical curvature is completed andcan be installed at the beginningof next year.

The design of the multianal-yser component for the multi-analyser/multidetector option ofPUMA is almost defined.

Figure 3.10:Photograph of the bending apparatus for the focussing analyser

The 11 PG(002) analyser crys-tals will be mounted on one com-mon table with individual rotat-ing devices for each analyserwhich again can be translated onindividual linear stages. Thus,the distance sample/analyser ofeach individual analyser can beadjusted. The electronics groupof the IPC has just finished aprototype of a multi stepper mo-tor control module for 6 mo-tors on a single card, thus pro-viding a highly integrated con-trol system for the sophisticatedmultianalyser system. Currently,the mechanical specifications ofthe multidetector unit will be de-veloped. Two different detec-tor options are planned: 11 sin-gle detectors, each individuallymovable and synchronized withthe corresponding analyser bladewill be used for dispersive modesof operation. An additional PSDdetector is

provided for the focussing mode.In cooperation with the FRM-IIdetector group discussions aboutthe detector electronic controlsystem for this option is ongoing.

In the course of this year mainparts of the instrument computernetwork were installed. Theclose cooperation with the IT-Network group of the FRM-II fordesigning, installing and testingmain parts of the instrument net-work is highly acknowledged.

The construction of the biolog-ical shielding around the instru-ment is carried out in coopera-tion with the mechanical designgroup of the FRM-II. It consistsmainly of heavy concrete-filledT-shaped steel containers (height2m, thickness 0.4m) which canflexible mounted and dismountedwith the help of the crane.

46 3 Spectrometers

3.5 RESEDA – Resonance Spinecho Spectrometer

M. Bleuel1, M. Axtner 1, R.Gähler2, P. Böni1

1 Physik Department E21, TU München2 Institut Laue-Langevin, Grenoble

Neutron guide

Figure 3.11:Scetch of RESEDA

RESEDA, the quasielasticneutron resonance spinecho(NRSE)-spectrometer at theFRM-II is located in the neu-tron guide hall. The neutronguide (5b) leading to the in-strument was accomplished in2003 and is the longest polarising

Figure 3.12:magnetic guide field for the polarising guide

neutron guide in the world. Be-cause of it´s length the bendingradius is big (r = 1640m), so thateven neutrons with a short wave-length (λ = 2Å) are transmittedto RESEDA. The polarisation isbetter than 96% over the wholewavelength range.

On the photo (3.12) the polaris-ing neutron guide in its magneticguide field is shown. A coverof 70 mm lead acts as a biologi-cal shield against theγ-radiation,which comes from absorption inthe boron glass.

Neutron selector

In 2003 a reconstruction of se-lector shielding became neces-sary, because a new polarised

reflectometer joined RESEDAand will use the excellent neu-trons from 5b-neutron guide aswell. During this reconstructionthe whole instrument, includingthe neutron guide and all sur-rounding installations, was plot-ted in a three-dimensional modelwith construction program "solidworks".

The new shielding is to be in-stalled in the early spring 2004.As soon the operation of theFRM-II has started, the final ad-justment of RESEDA with neu-trons can begin.

Activities in 2003

The main activities in year 2003at RESEDA were concentratedon the electric components ofthe instrument. The RF-circuitsand the detector electronics weretested in their final position,the encoder, which determinesthe position of the spectrome-ter arms, and the NSE coils forthe low resolution mode weremounted. Also the cabling ofthe instrument (motors, air cush-ions, air sensors, water, DC-and HF-supply for the Bootstrap-coils) was continued and theinstrument-control-program wascontinuously improved.

3 Spectrometers 47

Figure 3.13:Topview of RESEDAwith Selector Shielding

λ λ = 2,5−15flux at sample position Φ > 108/cm2s (λ = 4)spin echo time τ = 0,001−30nslength of the pseudo fields L = 2,6mmax. (pseudo) magnetic field B = 1200Gmax. momentum transfer (λ = 2.5) q = 3.5−1

Table 3.1:Characteristics of RESEDA

The estimated start of theFRM-II will not be before springof 2004. Because of this delay ofat least 3 years compared to the

early time table, all RESEDA ac-tivities are stretched. It is impor-tant to keep the know how andavoid a further splitting of theteam.

3.6 RSSM – Backscattering Spectrometer

P. Rottländer1, T. Kozielewski1, M. Prager 1, D. Richter 1

1 IFF - Forschungszentrum Jülich

The backscattering spectrome-ter RSSM will be built at theend position of neutron guideNL6. In its standard setup itwill offer an energy resolution of0.8 µeV and an energy windowof ±32µeV. A higher resolutionwill be optionally available. In2003, it made great advances to-wards completion next year. Inthe following, we will give ashort description of the instru-ment, together with an outline ofthe principal developments.

The first element which neu-trons encounter is the ELREBO,which is a device to position apiece of a neutron guide, a beamstop, or a liquid nitrogen cooledberyllium filter in the neutronbeam. The box itself is finishedand vacuum tested, and the beamstop will be finished by end of2003. The beryllium filter al-ready is in place but with too higha nitrogen consumption and a too

high final temperature of about140 K. A redesign is on the way.

After serious problems atthe University of Hannoverwith casting of the rotor ofthe home-made velocity se-lector which follows the EL-REBO, it was decided to buy

Figure 3.14:The intermediate version of the chopper wheel during center-ing of the central bushing

a commercial selector from As-trium, formerly Dornier. It willbe delivered in July 2004. Inthe wavelength range of about3 Å, this selector will be moreefficient which might alleviatethe cooling problems of the Befilter. However, the project of a

48 3 Spectrometers

home-made selector will not beabandoned, but adapted for theSi(311) option (see below).

The instrument specific conver-gent neutron guide has a separatevacuum chamber which was fin-ished last year. Due to problemson the side of the glass supplier,the guide elements will be deliv-ered in March 2004 only.

At a short distance after theneutron guide, the neutrons arereflected by the phase-spacetransformation (PST) chopper.It is divided into six segments,three of which carry crystalsof pyrolytic graphite at the cir-cumference, the three othersare empty. Neutrons comingfrom the chopper will be per-pendicularly reflected at themonochromator. On their wayback, they will pass the chop-per at a segment without crys-tal, and hit the sample. As amajor design feature of the spec-trometer, the chopper also in-creases the intensity of the neu-tron beam in the desired energyrange by accelerating neutrons

Figure 3.15:Test run of the final version of the Doppler drive at 4.7 m/s

which are too slow, and slowingdown those who are too fast. Formore information, see [1].

In the course of the year, thechopper wheel had to be com-pletely redesigned, primarily dueto safety reasons. The wheel nowconsists entirely of aluminum,where an alloy with high long-term fatigue limit has been used.Through successive design im-provements, mechanical stressduring rotation also has been re-duced. An intermediate (Fig.3.14) and the final version of thewheel both have been success-fully spun at the speed of 4850rpm; 4800 rpm will be the operat-ing frequency. The chopper driveis also assembled and has enteredthe test and optimisation phase.

The monochromator ismounted on a drive to achievethe energy variation needed forspectroscopy. A test version ofthe drive with an aluminum shafthas been successfully tested for amonth in the beginning of 2003.Using those experiences, the finalversion was built with more

built with more sophisticated airbearings which allow for a highermanufacturing tolerance. There-fore the shaft could be madeof carbon fibre compound and amore powerful motor employed.This gave an increase of themaximum attainable speed from3.5 m/s to 4.7 m/s. The finalversion ran successfully for oneweek in November 2003. The ap-proval of the drive will be in De-cember 2003.

After being scattered by thesample, neutrons of the rightwavelength are reflected, againperpendicularly, at the analysercrystals. Here, the instrumentwill offer a standard version ofunpolished Si(111) crystals, witha neutron energy of 2.08 meV,an energy range of±32µeV,and a resolution of 0.8µeV.It will also provide a high-resolution setup with polishedSi(111) crystals with a resolutionof 0.45µeV, a CaF2 monochro-mator which shifts the energywindow by 24µeV towards lowerenergies, and a version withSi(311) monochromator and anal-ysers. It uses neutrons with ahigher energy of incidence of7.63 meV. The latter has a lowerresolution but gives access toa different momentum transferrange.

The standard version of analy-sers and the monochromator mir-ror plate already were equippedwith Si(111) crystals in 2002. In2003, the first analyser shellswere equipped with polishedSi(111) crystals of 5×5 mm edgelength. In Fig. 3.16, a qualitycheck of the crystals is shown.This check is performed by plac-ing a laser close to the centre of

3 Spectrometers 49

Figure 3.16:A quality check of the high-resolution analyser. Left the veryfirst test waver that was glued, to the right a good average image.

the analyser sphere. The laserbeam is widened a little so that itilluminates an area of about 8 cmdiameter. The reflected light fallson a graph paper screen which isagain close to the centre. The re-flections generally fall into onesingle place which is not muchtaller than twice the crystal size,except of those of the misalignedcrystals.

In October 2003, the electronicdepartment of the Research Cen-tre Jülich finally started with thedesign of the instrument controlsystem. At the same time, the in-strument also started to leave firsttraces in the neutron guide hallin Munich: The plaster floor atthe instrument site has been re-moved to give place for a “danc-ing floor”, on which the heavycomponents – Doppler drive and

analysers – can be arranged fordifferent setups, and for the an-choring of the gas-tight instru-ment chamber (Fig.3.17). Thechamber itself is already fin-ished and will be assembled

Figure 3.17:The instrument chamber during an inspection in March

in Munich early in 2004.The instrument will also of-

fer a diffractometer option whichallows for a good angular reso-lution. It will be realised bya series of position sensitive de-tector tubes which cover scatter-ing angles from around 10◦ to140◦. This option is already con-structed and awaits to be manu-factured.

The standard version of the in-strument is expected to be opera-tional in the third quarter of 2004.The additional monochromatorsand the diffractometer option willfollow until end of the year.

[1] Schelten J., Alefeld B. In: Scherm R., Stiller H. H. (Editors),Proceedings of Workshop on NeutronScattering Instrumentation for SNQ, volume Report Jül-1954, FZ Jülich (1984).

50 3 Spectrometers

3.7 TOFTOF – Time of Flight Spectrometer Near Completion

T. Unruh 1, J. Ringe1, J. v. Loeben1, J. Neuhaus1, W. Petry 1,2

1 ZWE FRM-II, TU München2 Physik Department E13, TU München

Several milestones could bereached on the way to the com-pletion of the time–of–flight spec-trometer TOFTOF: The chop-per system has been installedand tested successfully. The605 detectors were tested indi-vidually including the preampli-fiers. The detectors and thepreamplifiers were mounted onthe racks and wired with thetime–of–flight electronics. Lastbut not least a stage with themeasuring cabin for the usersof the instrument has been builtup. The last main components of

Figure 3.18:Front view of the time–of–flight spectrometer at FRM–II withdisassembled shielding of the primary spectrometer (see text)

the spectrometer have now beendelivered: The radial collima-tor, the focussing neutron guideand the shielding for the pri-mary spectrometer. These com-ponents will finally be assembledin spring of 2004. Thus 2003has been a successful year for thegroup of the TOFTOF spectrome-ter at FRM–II.

A picture of the instru-ment is displayed in Fig.3.18.The shielding of the primaryspectrometer is disassembled.Thus the view is free tothe four vessels containing a

total of the 7 chopper discs. Itprovides synchronous and asyn-chronous operation of each disc(diameter: 600 mm) with speedsfrom about 100 rpm up to22000 rpm. The phase shift ofthe discs can be adjusted withan accuracy of about 0.05 degree.The speeds and phases of everydisc and the gear reduction ofthe frame overlap chopper can bechanged within the allowed lim-its during operation of the chop-persystem without further limita-tions.

The flight chamber of thesecondary spectrometer is posi-tioned at the end of the primaryspectrometer and is partly visi-ble behind the chopper vesselsin Fig. 3.18. This chamber alsohouses the racks with 605 de-tectors and the same number ofpreamplifiers. A cut–out of thedetector is pictured in Fig.3.19.On top of the flight chamberthe electronics rack can be seen<which includes the time–of–flightelectronics and some hardwarefor instrument control. The ca-beling has been finished for thewhole detector system which iscurrently being tested.

On the left hand side inFig. 3.18 two measuring cabinsare visible. The first one is forthe neighbouring REFSANS in-strument and the second one isfor the time–of–flight spectrome-ter. The TOFTOF group is look-ing forward to furnish the cabin

3 Spectrometers 51

and to control the first TOF exper-iments from there next year.

Figure 3.19:Photograph of a cut–out of the TOFTOF detector bank

52 4 Fundamental Physics

4 Fundamental Physics

4.1 MAFF — Munich Accelerator for Fission Fragments

D. Habs1,3, R. Krücken 2,3, M. Groß 1,3, W. Assmann1,3, L. Beck1,3, T. Faestermann2,3, R. Groß-mann1,3, S. Heinz1,3, Ph. Jüttner 4, O. Kester1,3, H.-J. Maier 1,3, P. Maier-Komor 2,3, F. Nebel2,3,M. Schumann1,3, J. Szerypo1,3, P. Thirolf 1,3, F.-L. Tralmer 4, E. Zech2,3

1 Ludwig-Maximilians-Universität, München2 Physik-Department E12, TU München3 Maier-Leibnitz-Labor f. Kern- und Teilchenphysik, Garching4 ZWE FRM-II, TU München

The Munich Accelerator forFission Fragments (MAFF) isa reactor-based Radioactive IonBeam (RIB) facility which is be-ing planned and set-up at theFRM-II mainly by the groupfor experimental nuclear physics(Prof. Habs, LMU) and thephysics department E12 (Prof.Krücken, TUM).

The goal of this installationis the production of very intensebeams of neutron-rich isotopesthat will be available for a broadrange of nuclear physics experi-ments as well as applied physicsand nuclear medicine. In a firststep the beam will have an energyof 30 keV (low-energy beam). Ina second step also a high-energybeamline will be set up wherethe ions are post-accelerated toenergies adjustable in the range3.7–5.9 MeV/u. An overview ofthe project can be found, e. g. in[1, 2]. Here we will summarizemainly the progress within thelast year.

Authorization procedure

As the FRM-II has received itsfinal operating license in May2003, the MAFF authorizationprocedure can commence. Somefinancial questions, however, stillneed to be clarified (total cost ofauthorization procedure).

Infrastructure

The feedthrough-plugbetween neutron guidetunnel and casemate

The neutron guide tunnel of theFRM-II will house the mechan-ics to move the target trolley andto operate the ion source changer.As the operational reliability ofthese components is crucial forMAFF, a manually driven backupsystem is considered necessary.Due to the high activity level inthe tunnel, this system must beoperated from outside. From thetechnical point of view, an ob-vious place is the neutron guidecasemate which is connected tothe neutron guide tunnel by open-

ings in the concrete wall thatwere before only used by the neu-tron guides with the remainingspace filled by shielding bricksin order to fulfill radioprotectionand security requirements.

We therefore placed a spe-cial plug containing severalfeedthroughs alongside a neu-tron guide in the wall opening.The feedthroughs will containshafts to operate the ion sourcetrolley in case of failure of theelectric drives along the beam-line in the neutron guide tunnelas well as gas pipes to connectthe cryopumps at the beamlineto the compressors that will beplaced in the casemate.

Figure 4.1:The feedthrough-plug be-tween neutron guide tunnel andcasemate.

4 Fundamental Physics 53

The ion source test bench

At the former hot chemistry labof the LMU technological labora-tory an ion source test bench isbeing set up.

A major (financial) part of in-stalling this test bench is thedecontamination and cleaning-upof the hot lab. Since the fundinghas finally been provided in sum-mer 2003, the work is meanwhilein progress. The old ventilationsystem has been dismantled and anew filter system is installed. Thenext step will be the removal ofthe contaminated glove boxes.

After decontamination of thelab, a mass separator which hasbeen made available by GSI willbe set up. With added diagnos-tics, this mass separator will bethe main device for testing ofdifferent target/ion source units.Also a connection will be madeto the next floor laser lab, as weare not only interested in surfaceion sources but also in laser ion-ization. Therefore lasers in thelaser lab will be used to operatea laser ion source. We will thenhave a full test bench availablefor ion sources before their use atthe FRM-II in in-pile position.

Moreover, a new glove boxand a welding station have beenpurchased and combined to aninert-gas welding station for

shower

entrance emergency exit

HV cageArgon box

magnet withchamber injection chamber

einzellension source

rack rack

rackdispersion chamber

collector chamber rack

magnet power supply

double rack

mirrormirror

laser beam

locker

locker

monitor

barrier

Figure 4.2:The planned layout ofthe ion source test lab.

handling radioactive materials.This station will also be placedin the ion source testing lab andwill allow the assembling of ionsources containing uranium andthe welding of the target con-tainer or other parts.

Development ofcomponents

The in-pile cryopanel

A considerable fraction of theradioactivity produced in theMAFF fission target will con-sist of short-lived gaseous activ-ity, like krypton or xenon fis-sion products or the halogensbromium and iodine. It isintended to prevent migrationand release of these volatile fis-sion products by localizing themon a He-cooled cryopanel untiltransformation into non-volatilespecies byβ decay. The re-quired operational temperaturebelow 20 K results from thevapour pressures of the gases andvapours that have to be pumped.At this temperature the usual restgas components (except hydro-gen) as well as the fission iso-topes of Kr, Xe, Br, I will befrozen at remaining vapour pres-sures in the UHV region (≤10−11 mbar). Such cryopanelswill be installed in the MAFFbeamtube at either side of the fis-sion source (SR6-a and SR6-b).Limited by the inner diameter ofthe beamtube (250 mm) as wellas by the MAFF fission source re-quired to be moved inside the cry-opanel, a compact design had tobe developed.

As an adequate solution a

Figure 4.3:Prototype of the MAFFcryopanel.

double-wall tube (gap width4 mm) with 6 spiral-like sepa-rated sections for in- and outletof the cold He gas was designed.The cryopanel length is 1 m, theinner diameter of the He double-tube amounts to 157 mm, theminimum value needed still to beable to move the fission sourceinside the cryopanel. The wallthickness of 2 mm is a compro-mise between the safety regula-tions for a pressure vessel on theone hand and the minimizationof the nuclear heating due to theneutron andγ flux. A passivefloating heat shield will reducethe thermal heat load form thesurrounding warm beam tube.Aluminum of the type 6061-T6was chosen as material for thecryopanel because of its cryo-genic specifications needed forthe operation at an input temper-ature of 10 K, while the outputtemperature will be about 20 K.In addition material specifica-tions and certificates for the usein a nuclear reactor environmentalready exist for Al 6061-T6.

The cryopanels will addition-ally provide the necessary high-vacuum conditions close to thefission source. The pumping ca-pacity of the cryopanels is de-pending on the pressure and thegas type, amounting to about12.7 l s−1 cm−2 for nitrogen and5.9 l s−1 cm−2 for xenon. The


overall pumping capacity of thecryopanels on each side of thebeam line will amount to about8 · 104 l s−1 for nitrogen and 4·104 l s−1 for xenon.

In order to build up the tech-nological knowledge for manu-facturing the cryopanel via cryo-genic shrinking and to allow forperformance tests, a prototypehas been produced by ACCEL(Bergisch Gladbach), which isshown in Fig.4.3.

Beam Transport

The MAFF beam transport sys-tem is guiding the radioactiveion beam from the fission source,through a mass separator to thebeam cooler. New simulationsusing the SIMION7 code, whichincludes fringe fields and higherorder effects have been con-ducted. Fig.4.4shows on the leftside a simulation result obtainedwith the COSY code to give anoverview of the beam line.

Within the simulations for thebeam transport to the mass sepa-rator several critical issues havebeen addressed:

1. The tube with the ion op-tical elements installed isbent downwards a few mmby its own weight.

2. There might be a transver-sal offset between thesource exit hole, and theextraction lens entrancehole.

3. Some neutron and gammaabsorbers must be imple-mented in the beam line.

4. High efficiency and lowemittance are required.

Solutions to those problemshave been found:

1. The bending of the tube iscounterbalanced by steer-ing voltages applied to oneof the quadrupoles withineach lens. Fig.4.4 showson the right the quadrupoleprototype.

2. To be able to deal withlarge offsets of up to5 mm, a segmented ex-traction electrode, with a15 mm hole is used. Volt-ages can be applied to allfour segments. In conjunc-tion with the steering elec-trodes in the first tripletthe offset can be correctedand the beam is focused onaxis.

3. Gamma and neutron ab-sorbing materials areplaced at the foci after thefirst and second triplet.

Triplet 1

Triplet 2

Doublet 1

Triplet 3

El. Deflector

Magnet 1

El. Deflector

Ion Source

Reactor Wall

Slit S

ystem

Figure 4.4:A COSY simulation of the beam line is shown on the left side.The picture on the right displays a prototype of the beam extractionquadrupole.

4. Optimisation of thequadrupol voltages lead to100 % transmission up tothe mass separator and anemittance of 12π mm mradat 30 keV.

The mass separator follows thelayout suggested by Mattauchand Herzog. It consist of an elec-trostatic and a magnetic dipolefor energy independent mass sep-aration. Two masses,m1 andm2

(with m1/m2 = 1.5) will be se-lected by the separator and trans-ported to different coolers. Thesimulation reveals a mass resolu-tion of approximately 170 for theseparator. The magnetic dipoleis designed in a way that allmasses will be focused in thesame plane. A slit system willbe installed there to stop the un-desired masses. If a deflectionvoltage is applied to the slits, thebeam can be bent in a way to in-crease the mass range


from m1/m2 > 1.2 to m1/m2 =1.8.

Ion beam cooler

Instead of applying the well-established technique of RadioFrequency Quadrupoles (RFQs)where the ions are transverselyconfined during the cooling pro-cess by a quadratic pseudopoten-tial we have investigated a newconcept where the RFQ structureis replaced by a stack of thinRadio Frequency ring electrodeswith varying inner diameters sur-rounding the buffer gas volume.Other than with RFQs the poten-tial created by the RF funnel isbox shaped which suggests thatspace charge effects limiting themaximum beam current in a RFQshould have less effect.

Following detailed MonteCarlo simulations, we have car-ried out first experimental inves-tigations on the RF funnel at theIGISOL facility in Jyväskylä(Finland) with beam currentsup to 20 nA. We found a verypromising behaviour of the trans-mission characteristics since thetransmission of 65 % proved tobe independent of the beam inten-sity over all the examined rangewhich revealed the applicabilityof the RF funnel as high-currentcooler. This observation can onone hand be explained by thefunnel electrode structure provid-ing a high angular acceptance atthe entrance and allowing for asmooth focusing of the ion beamduring the cooling process. Onthe other hand, the box shapedpseudo potential reduces ionlosses due to field distortionswhich concern especially parti-

cles on trajectories far off thecooler axis.

The use of RF funnels for ionguiding and focusing itself is notnew and has already been appliedsuccessfully in some fields ofphysics and chemistry. However,up to now no RF funnels havebeen applied for confining low-energy ion beams at buffer gaspressures around 0.1 mbar. Butnevertheless, an efficient coolingof intense low-energy beams isimportant for MAFF, where high-intensity high-quality beams arerequired.

24 m

29

m

Control Room

Miniball

1.5m 3.5m 3.15m 3.03m 2.1m

4m

3.4

6m

3.8

m3

.8m

11.1

m1.

45

m1.

45

m

Charge

Breeder

101MHz IH-RFQ101MHz IH-DTL

101

Mhz

Rebun

cher

20

2M

Hz I

H-D

TL

7 g

ap

IH

-Str

uctu

re

Low energy

area

Low energy

area

4.5

9m

BuncherChopper

Re-buncher

Nuclear

Spetroscopy

Chemistry

MAFFTRAP

Recoil

Separator

1m

MAFF EXPERIMENTAL HALL

Figure 4.5:Layout of the high-energy part of MAFF in the external experi-mental hall.

The MAFF accelerator

The high-energy part of MAFFaims at accelerating intense ra-dioactive ion beams of a mass-to-charge ratio< 6.5 up to energiesof 5.9 MeV/u. Design goals aretoprovide the experiments with ahigh quality beam both in longitu-dinal and in the transverse dimen-sion and to be able to tune finelythe energy of the beam from 3.6up to 5.9 MeV/u.

Another design goal is toachieve a very high transmission,in order to avoid high contamina-tion of the structure and hence an


easier maintenance of the acceler-ating structures.

The first part of the acceleratorchain consists of a charge breedersource [3] that delivers ions withA/q < 6.5 and in a 101.28 MHzIH-RFQ which brings the ions upto an energy of 300 keV/u.

Currently the design has beenmodified in order to accommo-date an external multi-harmonicsbuncher placed 3 meters up-stream the RFQ. This device willallow a time spacing of about80 ns between the main bunches,and it creates a bunch with a lon-gitudinal emittance smaller thanthe one that is formed inside theRFQ.

The matching section down-stream the RFQ consists of achopper which cleans the beamfrom the small satellite bunchesthat are formed inside the RFQ.

The following accelerat-ing steps are provided by a

101.28 MHZ IH-Tank that booststhe energy up to 1.45 MeV/uand three 202.56 MHz IH-Tankswhich bring the beam up tothe energies of 3.0, 4.2 and5.4 MeV/u respectively [4].There will be 2 additional shortIH structures with a limited num-ber of gaps which will be usedfor the fine energy variation.

The interval required for theenergy variation is between 3.6and 5.9 MeV/u so in the opera-tion regime the final energy willbe tuned from the last operatingtank which delivers a final energyclose to the required one and thetwo short cavities will be used inthe accelerating and deceleratingmode depending on the desiredenergy. In this way it is also pos-sible to vary the final longitudi-nal beam parameters to achievebeam focus at the target position.

The beam is then transportedto the high energy experimental

area by means of two identicalmodular beam lines. The designgoal of these beam lines are thetransport of the beam without dis-persion at the target point and themodularity, allowing hence a sav-ing of the power supplies and ofthe other services.

MAFF-Trap

A Penning trap system similar toREXTRAP at ISOLDE (CERN)or JYFLTRAP at Jyväskyläis being built to measure nu-clear masses with high precisionand/or to do nuclear spectroscopyat trapped ions. The supercon-ducting magnet has been orderedat the MAGNEX company andwill be delivered mid 2004. Thedevice will first be installed atthe low-energy beamline but canlater be moved to the high-energybeam.

[1] MAFF — Physics Case and Technical Description.http://www.ha.physik.uni-muenchen.de/maff/.

[2] Habs D., et al. The Munich Accelerator for Fission Fragments MAFF. Proc. EMIS-14, to bepublished in NIM B.

[3] Kester O., et al. In:Ann. Rep. of the EU-RTD project “charge breeding”, HPRI–CT–1999–50003(2002).

[4] Pasini M., et al. Radioactive ion beam acceleration. RNB 6 Conf. Proc., to be published.

4.2 MEPHISTO – the Cold Neutron Beam Facility for ParticlePhysics at FRM II

C. Schanzer1, O. Zimmer 1

1 Physik-Department, Technische Universität München, 85748 Garching

MEPHISTO is the cold neu-tron beam user facility for par-ticle physics experiments at theFRM II. Its experimental zoneis situated in the neutron guide

hall at the end of the 116× 50mm2 supermirror neutron guideNL3a. Some of the studies tobe performed at this facility werealready lined out in the last an-

nual report. Here we briefly de-scribe some of its features and theinstallation. The design of thebeamline enables an easy adap-tation to the actual experiment


Figure 4.6:Section of the guide NL3a which will transport cold neutrons toMEPHISTO. The support socket shown is situated six meters upstreamfrom the end of the guide. The glass elements can be easily reconfiguredto accommodate a neutron polariser and choppers or a velocity selectoron the socket.

without compromising a carefulshielding for low radiation back-ground.

The instrument can be oper-ated either with polarised or non-polarised neutrons. A large-area,40 cm long supermirror benderpolariser (manufactured by SwissNeutronics) is already available.The device possesses coatingswhich stay remanent even at in-verted magnetic guiding field,and it is able to polarise thewhole white beam. It may beinstalled on a support socket inthe casemate six meters upstreamfrom the end of NL3a, thus tak-ing advantage of the heavy shield-ing of the casemate’s wall visi-ble on the right side in Fig.4.6.Movable support structures andshieldings enable a reasonablysimple adaptation to the changeof the beam direction

by about 20 mrad induced by thepolariser.

A secondary beam shutter issituated in the vacuum of the

Figure 4.7:The secondary beam shutter (technical design: F. Tralmer). Itsvacuum housing also includes two movable beam attenuators, visible be-hind the shutter block.

beamline 10 m upstream from theend of the neutron guide (see Fig.4.7 and4.8). It shall enable theuser to work in the experimentalzone at running reactor withoutinterfering with the other instru-ments. The end of the neutronguide is equipped with a flangesuch that the whole beamline in-cluding the tank of the beamstopcan be configured to consist ofa single vacuum, thus avoidingneutron windows with their inher-ent background due to neutronabsorption and scattering. Forexperiments requiring ultra-highvacuum, two special flanges withvery thin foils may be used in-stead, enabling a separation ofvacua of different quality.

The heavy beam stop includesseveral degrees of freedom fora quick adaptation to the actualparticle physics experiment, with-out any compromise concerningsafety requirements. Neutron ab-sorbing materials and


and lead shieldings for gamma-suppression are situated in a 2m long vacuum tank, which isequipped with a CF-flange forits connection to the beamline.In a later stage, the tank willalso house a little supermirror po-lariser and a neutron detector formonitoring the stability of theneutron polarisation during ex-periments without breaking thevacuum.

Figure 4.8: J. Fink mounting theshutter in the NL3a beamline.

Figure 4.9:The team of the central workshop with their readily welded vac-uum tank of the beamstop, a fine piece of craftsmanship. From rightto left: K. Hochstetter, A. Braunschedel, T. Fechter, R. Jahrstorfer, M.Reither and M. Pfaller (M. Ducke and M. Schönberger missing on thephoto).

4.3 NEPOMUC – Neutron Induced Positron Source Munich andits Experimental Facilities

C. Hugenschmidt1, G. Kögel2, R. Repper1, K. Schreckenbach1, P. Sperr2, B. Straßer1, W. Trift-shäuser2

1 ZWE FRM II, TU München2 Institut für Nukleare Festkörperphysik, Universität der Bundeswehr München, 85577 Neubiberg

NEPOMUC

The intense positron source is in-stalled as an in-pile componentin the tip of the declined beamtube SR11 at the new researchreactor FRM-II at the TechnicalUniversity of Munich. After ther-

mal neutron capture in cadmiumthe absorption of the releasedhigh-energyγ-radiation in plat-inum generates positrons by pairproduction. At NEPOMUC theworld highest intensity of the or-der of 1010 slow positrons per sec-ond in the primary beam is ex-

pected. The moderated positronsare extracted by electrical lensesand magnetically guided to theoutside of the biological shield-ing of the reactor. There, an aper-ture, gamma-detectors and a mov-able beam monitor are installedin order to measure the


Figure 4.10:Overview of NEPOMUC and its instrumentation.

positron intensity as well as thebeam profile. The followingbeam line guides the positrons toa beam splitter on the experimen-tal platform.

INSTRUMENTS ATNEPOMUC

In the first arrangement, thepositron beam is connected tothree experiments on the plat-form in the experimental hallof FRM II (see 4.10). Withinour collaboration apulsed lowenergy positron beam system(PLEPS) and ascanningpositronmicroscope (SPM) will be trans-ferred to NEPOMUC. Both,PLEPS (4.11) and SPM (4.12)were constructed and built at theuniversity of the Bundeswehr inMunich (UniBW). At present,these facilities are under

operation with 22Na positronsources with beam intensities upto 103 positrons per second. Inorder to get depth dependent in-formation from the surface to thebulk of the material, the positronenergy can be varied from a few100 eV up to 18 keV. The lat-eral resolution is about 3 mm atPLEPS and of about 2µm at themicroscope. The third instrumentis a facility for positron anni-hilation inducedAuger electronspectroscopy (PAES). The PAESfacility was built up at E21/TUMand is designed for extremely sen-sitive studies in surface science(4.13). The Auger process isusually initiated by photo- or im-pact ionisation of core electronsby irradiation with X-rays or keV-electrons. An alternative tech-nique to induce core shell ioni-sation is the annihilation of core

electrons with slow positrons.Due to the low positron energyof some 10 eV, no background ofsecondary electrons is producedin the higher energy range ofreleased Auger electrons. Be-sides this advantage, PAES isan exceptionally surface sensitivetechnique since most of the im-planted positrons annihilate withelectrons of the topmost atomiclayer. The positron Auger elec-tron spectrometer was extendedby a separate sample lock and asample preparation chamber. Inorder to prepare clean and well-defined surfaces, it is equippedwith an argon sputter source andan adapted sample heater for in-situ annealing of defects partic-ularly in the near surface re-gion. An effusion cell was in-stalled which consists of a tung-sten foil boat with tunable sampledistance. In addition, an electronbeam evaporator was mountedfor evaporation of low vapor pres-sure materials. Both, the effusioncell as well as the electron beamevaporator allow accurate depo-sition of adsorbates in the sub-monolayer range. The exact layerthickness of deposited material isdetermined by a piezo oscillator

Figure 4.11: Pulsed low energypositron beam system (PLEPS)under operation at UniBW.


crystal. After preparation, thesample is transferred through agate valve into the analysis cham-ber by a magnetic coupled rod.

Figure 4.12:The scanning positronmicroscope (SPM) under opera-tion at UniBW.

Figure 4.13:Facility for positron annihilation induced Auger electron spec-troscopy (PAES) at E21, TUM.

5 Irradiation and Radiography Facilities 61

5 Irradiation and Radiography Facilities

5.1 ANTARES – Construction of the Neutron Radiography andTomography Facility

B. Schillinger1,2, E. Calzada1,2, F. Grünauer 1,2

1 ZWE FRM-II, TU München2 Physik Department E21, TU München

Overview

In 2003, the components for theneutron radiography and tomog-raphy facility ANTARES wereconstructed and assembled atbeam line SR4b in the north-east corner of the experimentalhall. The design of ANTAREShas been explained in much de-tail in the previous annual report,so only a short overview is givenhere before details of the assem-bly are described. ANTARESconsists of an external beamshutter, flight tube and block

Figure 5.1:Complete layout of the facility: 1. Collimator inside of the bio-logical shielding, 2. External vertical beam shutter, 3. Iris diaphragm, 4.Pneumatic shutter, 5. Flight tube, 6. Variable beam size limiter, 7. Mea-surement cabin, 8. Wall and beam stop, 9. Detector and camera system,10. Sample manipulator, 11. Sliding door, 12. Flight tube shielding,

house. The external beam shut-ter is necessary because the UCNexperiment at SR4a will insert anozzle through the drum shutterin the biological shielding, ren-dering it unusable for ANTARES.Fig. 5.1shows an overview of thefacility.

Construction of the wallelements

The shielding elements and blockhouse walls were constructed assteel casings filled with heavyconcrete. The secondary shutter

of density 4.5 housing and theblock house elements were filledwith concrete tons/m3. As thecollimated beam inside the flighttube does not touch the walls, adensity of 3.5 tons/m3 was suffi-cient for the walls of the flighttube housing. As the crane inthe experimental hall has a nom-inal capacity of 10 tons only,all elements were designed fora maximum weight of about 9.5tons. For some elements of theblockhouse wall near the beamentrance, this weight constraintcould not be met, so additional re-movable concrete filled cylinderswere constructed inside the wallvolume , which were mounted af-ter the assembly of the wall ele-ments.

Additional problems arose be-cause the concrete company guar-anteed only for the minimum den-sity of the concrete, but not forthe nominal density. In reality,the density became higher thanthe calculated values, leading toa component weight of more thanten tons. Fortunately, the reactorhall crane has a special 10% over-load range (of which the use is re-stricted to a

62 5 Irradiation and Radiography Facilities

Figure 5.2:Moving the wall with aircushions

few hours per year), which wascapable of handling all elements.

Mounting of the wallelements

The blockhouse walls weremounted from five horizontal seg-ments each, weighing more than50 tons each. The assembledwalls were fitted with mountingframes for up to six air cushions(Fig. 5.2) . An external com-pressor was used to lift the wallsand to move them into place, asthe capacity of the house installa-tions was not sufficient.

Unfortunately, several cablechannels in the hall floor wereonly roughly covered with un-even steel plates, rendering an un-even surface and an escape pathfor the compressed air. Theyhad to be covered with thin steelplates, but still, the walls couldonly transit the uneven regionswith the aid of two steel cabletackles affixed to the reactor hallwalls (Fig.5.2).

The flight tube

As described in the last report,the flight tube consists of individ-ual segments connected by rub-ber sealant rings, mounted on C-

shaped brackets. The original de-sign was taken from the vacuumhousings for the neutron guidesleading to the guide hall. Nospace was available top mountcircular flanges, so the originalsquare shape with rounded cor-ners was altered by roundingthe straight sections to a convexshape (Fig.5.3).

Each segment has an individ-ual vacuum connector, so singlesegments can be removed for ex-perimental installations, the re-maining segments can be closedoff by covers. Vacuum pumpand vacuum control system weresalvaged from the old reactor.For the final neutron window atthe end of the flight tube, amassive aluminum cover with athin-milled window was foreseen.Static calculations predicted amaterial failure at the edge of themilled area, so the cover was con-structed as a sandwich of a 3 mmaluminum plate, rubber seals andflange rings.

The cover stones for the flighttube housing have retractableball casters on which they canslide from the middle, crane-accessible position to their finalposition under the platforms.

Figure 5.3:The flight tube with con-vex flanges

Mounting of theblockhouse roof

The area of the blockhouse isnot accessible by the hall crane.To be able to mount the roofat all, it was constructed fromsmall two-ton bars. The orig-inal intention had been to usea balance bar and counterweightwith the hall crane, which is ableto telescope towards the cornerof the hall just to the beginningof the block house. Later on,a special compact truck-mountedcrane was found which was justable to drive into the reactor hallin the time when the other experi-ments at SR-2 and SR-3 were notyet assembled, with 5 cm spaceleft next to the already mountedmeasuring cabins. The roof ele-ments were deposited on the plat-form for the positron experimentwith the hall crane, the truck-mounted crane took over fromthere and deposited them in posi-tion on the blockhouse roof (Fig.5.4).

The external beamshutter

The external shutter was con-structed as a vertical concreteblock containing two differentcollimators, driven by a hy-draulic piston (Fig. 5.5) . Aservo system with a position en-coder and a control valve can po-sition the shutter block to the ac-curacy of a tenth of a millime-ter to a collimator position or tothe closed position. When the po-sition is reached, two hydraulicclamps hold the block in position,the hydraulic pump shuts off.


Figure 5.4: Mounting the block-house roof elements with a truck-mounted crane

A hydraulic pressure tank holdssufficient pressure to open the hy-draulic clamps and close the shut-ter within one third of a second

when magnetic valves are openedin case of power failure.

Outlook

In December 2003, The electricinstallation for the blockhousewas completed, so we had "firstlight" in the blockhouse evenwithout neutrons. In 2004, thewalls will be covered with B4Crubber mats to reduce the ac-tivation of the steel walls, thesample manipulator will be trans-ported to the blockhouse and thedetector box will be completed.The fast experimental pneumatic

shutter and the filter wheel forphase contrast tomography willbe mounted as the final installa-tions, expecting the first measure-ments soon after commissioningof FRM-II.

Figure 5.5:The hydraulic servo sys-tem for the beam shutter

5.2 NECTAR

T. Bücherl 1, Ch. Lierse von Gostomski1, E. Kutlar 1, E. Calzada2

1 Institut für Radiochemie, TU München2 ZWE FRM-II, TU München

General

The NEutron Computer Tomog-raphy And Radiography (NEC-TAR) facility is set up at the con-verter facility in the experimen-tal hall (beam tube SR 10). Itsuniqueness is the use of fissionneutrons to radiograph samplesof volumes up to 0.5 m3 and 400kg maximum burden. The fastneutron flux will be up to 6.4·107

cm−2 s−1 and the collimation isadjustable to L/D values of up to300.

Current status

The main components in the mea-suring room are the manipulatorfor the objects, the manipulatorand storage system for the

Figure 5.6:View in the measuring room with manipulator system for theobjects (left), for the detectors (right) and provisionally mounted two-dimensional detector system (without CCD-camera).


detectors, the actually two detec-tor system themselves and the re-movable and changeable collima-tor set in the partition wall to theneighbouring room for medicalapplications. All these compo-nents must be adjusted preciselyto the incoming neutron beam.As most of them are heavy inweight, this cannot be performedmanually within the accuracy re-quired. Therefore a crane to bemounted on the ceiling was de-signed and its statics calculated.The latter is also required for itslicensing.

Although the crane is not in-stalled yet, both manipulatorshave been set up provisionally toverify the correct layout of thearrangement (Fig. 5.6) and todesign appropriate clampings forthe detector systems. To be moreflexible in positioning the objectmanipulator, it was additionallyequipped with a connection box,i.e. all cables coming from andgoing to the control unit are con-nected with the manipulator viathis box.

Figure 5.7:Control unit for the ma-nipulators.

The functionality of the con-trol unit (Fig. 5.7) was ex-tended by a three phase trans-former, thus avoiding problemsshowing up when connected toelectronic circuits having FI pro-tective switches. To avoid un-necessary (neutron) activation ofthe object and the equipment, theeffective beam cross-section willbe limited by an adaptable colli-mator system to the real measure-ment geometry, i.e. the dimen-sions of the object and/or the ac-tive detection area. It will con-sist of a number of boxes madeof iron which fit one into theother. The layout is shown inFig. 5.8. The development of theLABView-based software pack-age for control of the manipula-tors and detectors was continued.

Next steps

The work in 2004 will be sepa-rated into three phases. In thefirst phase the set up of the hard-ware will be completed, e.g. thecrane, the detector systems, thecontrol cabin and the collimatorswill be installed and all compo-nents adjusted exactly. The com-pletion and optimization of thecontrol software package and itsintegration with all other soft-ware modules (e.g. those of thedetector and reconstruction soft-ware packages) is the main partof the second phase. At the endof phase two, the system is readyfor operation. With the availabil-ity of fission neutrons the thirdphase will start. Extensive testsof the detector systems will de-fine their characteristic parame-ters like the absolute detectionefficiency, the spatial resolution(for the two-dimensional position

Figure 5.8: Layout of the adaptablecollimator system, which will beplaced in the partition wall.

sensitive detector system), thegamma-sensitivity and theneutron-gamma-discriminationratio, respectively etc. Basedon these results possibilities forfurther optimization will be de-termined and – if possible – real-ized.

As NECTAR is an unique sys-tem, no extensive a-priori infor-mation on its probable perfor-mance is available, i.e. the in-fluence of beam hardening, for-ward scattering in the sample etc.on the resulting data for differ-ent materials must be determinedfirst. This will be performed byusing different test objects, likestep wedges of many differentmaterials. Based on this knowl-edge, the basics for a successfulevaluation of tomographic datawill be set. Tomography of manyobjects, different in geometry, di-mension and material composi-tion will complete the third phaseand indicate the starting point forroutine and research operation.


5.3 Irradiation Facilities

H. Gerstenberg1, X. Li 1


The FRM-II will be equippedwith various irradiation facilities.Due to the different tolerablesample volumes, neutron fluxdensities and obtainable neutronfluencies they allow in their en-tirety a broad scope of possibleapplications in basic and appliedscience.

In particular during the nuclearstart-up phase of the FRM-II thefollowing facilities will be avail-able:

• a pneumatic rabbit system(RPA),

• a hydraulic rabbit system(KPA),

• a test-rig for a silicon dop-ing facility (SDA),

Figure 5.9:Control and handling room of the pneumatic rabbit system. Left:transfer pipe switches, middle: sample handling box, right: sample load-ing stations.

• a pneumatic rabbit system(TRP) for prompt transportof irradiated samples intothe neighbouring institutefor radiochemistry.

In addition, at least a highspeed pneumatic rabbit systemwill be added in the future.

All of the irradiation positionsare located within the heavy wa-ter moderator tank and can beloaded and unloaded during re-actor operation. A major advan-tage of these irradiation positionsis the availability of a very purethermal neutron field. Thus par-asite threshold reactions by fastneutrons and the production oflarge clusters of irradiation de-fects are suppressed effectively.

The pneumatic rabbit system(RPA) serves typically for the

activation of samples for neutronactivation analysis. The pneu-matic devices are operated by car-bon dioxide gas in order to avoidthe production of radioactive Ar-41 during the irradiation of air.The facility is composed of 6independent irradiation channelsbeing identical except for the ver-tical position within the modera-tor tank. Consequently the neu-tron flux densities in the positionsvary between 5·102cm−2s−1 and2 · 1014cm−2s−1 and the appro-priate irradiation position can bechosen to fit the individual task.For radiation protection reasonsthe entire handling takes place ina room being exclusively used forthis facility. Irradiation times canbe chosen between about 20s ,i.e. a significantly longer timecompared to the transport time ofthe rabbits of a few seconds, and5 hours. The maximum irradi-ation time is limited by the useof polyethylene rabbits as outerirradiation package. After irra-diation the sample is first trans-ported into a shielded cooling po-sition where the induced gammadose rate is measured. Finallyit is transported back into thehandling room (Fig.5.9) whereit is either removed from the ir-radiation facility by remote han-dling tools or alternatively sentdirectly to the institute for radio-chemistry via a connected rabbitsystem (TRP).


Figure 5.10:interactive control screen of the hydraulic rabbit system (KBA).The current status is displayed directly. The transfer system can be oper-ated automatically and manually.

In contrast to the pneumaticsystem the hydraulic rabbit sys-tem (KBA) will be used for longterm exposure or the irradiationof larger sample volumes up toapproximately 35 cm3. Con-sequently the samples are con-tained in Al rabbits during irra-diation, which however are per-forated in order to guarantee thecooling of the samples by poolwater during irradiation. If nec-essary the direct contact betweensample and pool water is avoidedby the use of sealed quartz am-poules or small welded inner Alcontainers. The facility offers 2independent identical irradiationchannels each of which can beloaded by up to 5 rabbits at thesame time. The driving mediumof the facility is reactor pool wa-ter. The neutron flux densities

have been calculated to be about4 · 1014cm−2s−1. Since ratherhigh activities will have to be han-dled in the hydraulic rabbit sys-tem, most of the handling proce-dures in particular the receipt ofirradiated samples and the mea-surement of the induced gammadose rate is done remotely underwater about 3.5 m below the re-actor pool level. In addition alsothe loading of the samples intolead shielded transport containerstakes place under water.

Both rabbit systems are con-trolled automatically by the SPScontrol system through a PC withWindowsNT operating system re-spectively. The current situationof each transfer system, includ-ing the position of the irradiationcapsules and the status of all con-trol valves and other important

parameters such as system pres-sure and gamma dose rate at thesample cooling positions can bedirectly displayed on the screenof the control PC (Fig. 5.10).If necessary, a semi automaticcontrol through direct operationof single components by mouseclick is also possible.

The silicon doping facility(SDA) will serve for the P dop-ing of the semiconductor Si bymeans of the nuclear reaction30Si

(n,γ)−−→ 31Si(β−)−−→ 31P.

This so called neutron trans-mutation doping offers the mosthomogeneous doping profile be-ing necessary particularly in highpower electronic components. Inorder to fulfil the requirementof homogeneity, however, it isof crucial importance to guaran-tee a constant neutron flux den-sity over the entire volume ofthe Si ingot. Therefore the in-got is rotated during irradiationand the irradiation position hasto be equipped with a speciallyshaped absorber (Fig.5.11). Dur-ing the nuclear start-up phase theneutron flux distribution in theSi ingot will be determined ex-perimentally by using a simpli-fied test rig. Then, based onthis flux distribution the exactshape of this absorber made fromNi can be calculated and con-structed. Once the absorber pro-file will be known the final Sidoping facility offering a semiautomatic operation will be com-pleted. It will be suitable for Si in-gots with diameters of up to 200mm and a capacity of several tonsper year.

Till the end of 2003 the hy-draulic rabbit system (KBA) andthe test rig of the silicon doping


system (SDA) are complete. Astart-up test on them has alreadybeen carried out successfully be-fore the commissioning of the re-actor. The pneumatic rabbit sys-tem (RPA/TRP) is already in thelast phase of its start-up test.

Samples with different weightand irradiation time are loadedand transferred during this start-up test of the rabbit systems un-der all possible conditions in or-der to simulate the real situa-tions. Tests of several irradia-tion tasks in the same time andsame transfer system are also per-formed. The measurements ofgamma dose rate at the samplecooling positions are simulatedby using a conventional calibra-tion gamma source. If the al-lowed limit is overshot, the sam-ple transfer is prevented by a in-terlock at the cooling positionsdue to the aspects of the irradia-tion protection.

At the test-rig of the Si dop-ing system loading and transportof Si crystals between the irradi-ation position in the moderatortank and the loading place in thestorage pool are tested. Flux mon-itors (Al, Au, Zr etc.) for thedetermination of the neutron flux

distribution inside a Si ingot arealready prepared. They will beplaced at a total of 18 positionswithin one Si ingot during the nu-clear startup.

Figure 5.11:scheme of the silicon doping system.

68 6 New Developments

6 New Developments

6.1 Novel Type of Neutron Image Plates Based on KCl:Eu2+-LiF

M. Schlapp1, M. Hölzel 1, R. Gilles1, A. Ioffe 3, T. Brückel 3, H. von Seggern2, H. Fuess2

1 ZWE FRM-II, TU München/TU Darmstadt2 Fachbereich Materialwissenschaft, Technische Universität Darmstadt3 Inst. f. Festkörperforschung, Forschungszentrum Jülich

Introduction

Neutron image plates (NIPs)provide a mean for the two-dimensional, position-sensitivedetection of neutrons. They com-bine the advantages of a largedynamic range in dose (5 ordersof magnitude), a high spatial res-olution (<200 µm) and a highefficiency for the detection ofneutrons [1]. NIPs have foundwidespread application as neu-tron detectors for single-crystaland powder diffraction, small-angle scattering and tomography.After neutron exposure, the im-age plate can be read-out by scan-ning with a laser. Commerciallyavailable neutron image plates(NIPs) consist of a powder mix-ture of BaFBr:Eu2+ and Gd2O3

dispersed in a polymer matrixand supported by a flexible poly-mer sheet. Since BaFBr:Eu2+ isan excellent X-ray storage phos-phor, these NIPs are prone toabsorb γ-radiation which is al-ways present as a backgroundradiation in neutron experiments.Several ways have been proposedto make X-ray image plates sen-sitive to neutrons: while severalgroups investigated the combi-

nation of BaFBr:Eu2+ as phos-phor and Gd2O3 as converter[1, 2], fewer reports were pub-lished on a possible usage of6LiF as converter [3]. Recently,we were able to demonstratethat a combination of KCl:Eu2+

as phosphor and6LiF as con-verter shows a promising poten-tial for neutron image plates witha low γ-sensitivity [4]. In thispresentation we report on theresolution of ceramic NIPs (C-NIPs) based on KCl:Eu2+ andLiF which were fabricated intoceramic image plates in whichthe alkali halides are acting asa self-supporting matrix withoutthe necessity for using a poly-meric binder. A simulation ofthe applicability of these NIPsas well as commercial platesfor the SANS apparatus at theinstrument Structure POwderDI ffractometer (SPODI) at theFRM-II will also be given.

Experimental

Ceramic neutron image plates (C-NIPs) were fabricated by press-ing a mixture of 30 mol%KCl:Eu2+ and 70 mol% LiFwith 2.5 wt% of a pressing

additive (Poly-[vinyl butaryl-co-vinyl alcohol-co-vinyl acetate])into pellets with thicknesses rang-ing from 250 µm to 1500 µmusing uniaxial and isostatic pres-sure of up to 5 GPa. In or-der to determine the spatial res-olution of these image plates,they were exposed to neutronsat the general purpose cold neu-tron beam at ForschungszentrumJülich. The sharp edge of aGd-metal sheet was used as amask. A numerical differentia-tion of a scan over this edge fol-lowed by a Fast Fourier Trans-formation yields the ModulationTransfer Function (MTF), a mea-sure for the contrast transfer ofthe original image onto the mea-sured signal. In general, theMTF is decreasing with an in-creasing spatial resolution givenin line pairs per mm (lp/mm), i.e.the contrast transfer from the im-age to the signal is diminishingwith finer details. Setting theminimum contrast that two pix-els must have to be distinguish-able to 20%, one can determinethe spatial resolution from theMTF. The newStructurePOwderDI ffractometer SPODI currentlyunder construction at the FRM-

6 New Developments 69

II reactor in Garching is goingto be equipped with a SANSapparatus [5]. Simulations onthe properties of the instrumentSPODI as well as the achievableQ-vectors of the SANS apparatuswere done with the software pack-age McStas of Risø National Lab-oratory.

Results and Discussion

The resolution of an image plateis determined less by the size ofthe read-out laser beam than bythe scattering of this laser lightwithin the image plate. Due todifferences in the indices of re-fraction of the materials used forthe fabrication of the NIP, thefocused laser light is smearedover a pear-shaped volume andall information contained withinthis volume contributes to thedetected signal proportionally tothe light intensity. It is obviousthat the lateral spreading of stim-ulation light within the NIP in-creases as the thickness of the im-age plate is increased since thephotons are scattered back andforth within the plate while inthinner NIPs they are emitted ei-ther at the front or back side. InFig.6.1 the experimentally deter-mined MTFs of ceramic neutronimage plates are depicted. Onecan clearly observe that the MTFis declining faster for thickerplates, i.e. the spatial resolu-tion is decreasing with increasingthickness of the plate. It can fur-ther be found that the reductionin the resolution is largest, whenthe overall thickness of the plateis small while for thicker platesthe effect of a loss of resolution isless pronounced. This can be at-

tributed to the fact that althoughthe lateral scattering of the stim-ulating light increases with thethickness of the plate, the inten-sity of the light is reduced byabsorption. Since the intensityof the detected signal is directlyproportional to the stimulating in-tensity, the light which is widelyspread laterally causes only aweak contribution to the signal.

In order to ensure a highdetective quantum efficiency,nearly all neutrons should beabsorbed in the NIP consist-ing of KCl:Eu2+ and LiF. Toachieve a high absorption andto stimulate all defects withinthe image plate, the thicknesshas to be in the range of 1000µm. Using a contrast trans-fer of 20 % as a measure, oneobtains a resolution of 1.5 linepairs/mm, i.e. a minimum pixelsize of 670µm. Although thisvalue is much larger than the

Figure 6.1:MTFs of ceramic NIPs with various thickness, setting the min-imum contrast that two pixel must have to be distinguishable to 20 %(MTF=0.2), one can determine the spatial resolution of the NIP from theMTF-curve

200 µm resolution which canbe achieved with a 135µmthick, commercially availableNIP (Fuji BAS ND) based onBaFBr1−xIx:Eu2+ and Gd2O3

[1, 6], one has to keep in mindthat the NIPs based on KCl:Eu2+

and LiF exhibit a much smallerγ-sensitivity and were devel-oped for experiments in a highγ-environment. While C-NIPsbased on KCl:Eu2+ and LiFshowed a detective quantum effi-ciency comparable to a Fuji BASND image plate, the signal afterexposure to 558 keV and 618keV γ-rays from a (n,γ)-reactionat cadmium was by one order ofmagnitude lower.

Due to the long distance of5 m between the monochro-mator and the sample posi-tion and small (2-4 mm diame-ter) pin-hole slits, a low diver-gence of the neutron beam atSPODI is achievable. Since this


instrument uses thermal neutronswith a wavelength between 1 and2.5 Å, the resulting scattering an-gles are small compared to SANSmachines using cold neutrons.This smaller angles are adaptedby using a NIP as detector whichhas a much smaller pixel sizethan the detectors normally usedat cold-neutron SANS machines.Fig. 6.2 exhibits the calculatedmaximum and minimum scatter-ing vectors Q as well as the res-olution in Q achievable with acommercially available and a ce-ramic neutron image plate (C-NIP) based on KCl:Eu2+ andLiF. Although the resolution ofC-NIPs is less, it is still well be-low the minimum scattering vec-tor achievable with this instru-ment and it can be expected thatthe usage of C-NIPs places lessdemand on shielding of the de-tector against a application of theNIP is the characterisation of theprimary neutron beam shape andγ-background. Additionally, for

Figure 6.2:Q-range covered by the SANS apparatus at SPODI

short sample-detector distances,the NIP renders the detection offull Debye-Scherrer cones possi-ble allowing a determination ofthe sample crystallinity and tex-ture. This results show, thatin principle particles with sizesranging from 1 to 900 Å can be

investigated by a compact SANSapparatus for thermal neutronsutilizing image plates as detec-tor. Using a larger wavelength- a second monochromator is un-der planning - the achievable min-imum scattering vector Qmin canbe reduced even further.

[1] Karasawa Y., Niimura N., Tanaka I., Miyahara J., Takahashi K., Saito H., Tsuruno A., Matsub-ayashi M.Physica B, 213&214, (1995), 978.

[2] Buecherl T., Rausch C., von Seggern H.Nucl. Inst. & Meth. A, 333, (1993), 502.[3] Takahashi K., Tazaki S., Miyahara J., Karasawa Y., Niimura N.Nucl. Inst. & Meth. A, 119, (1996),

502.[4] Schlapp M., von Seggern H., Massalovich S., Ioffe A., Conrad H., Brueckel T.Appl. Phys. A 74

(Suppl.), 74, (2002), S109.[5] Gilles R., Krimmer B., Saroun J., Boysen H., Fuess H.Mat. Sci. Forum., 378-381, (2001), 282.[6] Tazaki S., Neriishi K., Takahashi K., Etoh M., Karasawa Y., Kumazawa S., Niimura N.Nucl. Inst.

& Meth. A, 424, (1999), 20.


6.2 Response Function of a Neutron Detector for RadiationProtection Purposes

F. Grünauer 1

1 Physikdepartments E21, TU München

At FRM-II commercially avail-able neutron detectors will beused for radiation monitoring andprotection. These detectors areapproved in the energy range ofthermal to fast neutrons. Con-trary to usual applications ofneutron dosimetry (e.g. reactorhalls), there has to be expecteda major contribution of cold neu-

trons in the neutron-guide hallof FRM-II. It had to be exam-ined whether this detector is suit-able for cold spectra and if so,what calibration and security fac-tors have to be applied underthese special conditions. MonteCarlo calculations were carriedout in order to assess the responsefunction with respect to neutron

energy and angle of incidence.It was shown, that the responsefunction below thermal energiesis rather flat. Therefore the de-tector can be used even in thesubthermal energy range. The re-sponse of the detector is very de-pendent on the angle of incidenceof radiation.

Figure 6.3:A) Monte Carlo model of the neutron detector, B) vertical cut through the detector, C) horizontal cutthrough the detector, D) response function of the detector


6.3 Image Deconvolution in Neutron Radiography

F. Grünauer 1


Nearly all neutron radiographyfacilities use a setup similar toa pinhole camera. Lens systemsafford a very high constructioneffort. Generic pinhole systemsprovide high resolution at bigL/D ratios, whereL is the dis-tance between aperture and spec-imen andD the diameter of a cir-cular aperture. As the neutronfluence decreases proportional to(D/L)2 a high resolution meanslong exposure time and a badsignal to noise ratio. It wouldbe desirable to record imageswith high intensity and low res-olution and to correct the result-ing blur by an appropriate algo-rithm. Deblurring of images wastested with aL/D = 50 projec-tion (Fig. 6.4 B) of a test speci-men (Fig. 6.4 A), that was sim-ulated by Monte Carlo Method.The projection is the convolutionof the point spread function bythe ’ideal image’ (in this context,the ideal image is the projection,produced by a point source). Con-volution in the spatial domain cor-responds to a multiplication inthe fourier domain. In princi-ple the ideal image can be recon-structed by dividing the fouriertransform of the projection bythe fourier transform of the pointspread function and back trans-forming the result to the spatialdomain. However this ’invers fil-tering’ is

Figure 6.4: A) Test specimen, B) unprocessed projection, C) reconstruc-tion by a Wiener filter, D) reconstruction by van Cittert algorithm, E)reconstruction by Richardson-Lucy method

not successful if the image con-tains noise: Division by small ab-solute values of the point domaincauses a big damage in the recon-structed image. This effect canbe reduced by a Wiener filter [1],that gives a certain weight to eachpoint in the fourier domain dur-ing the filtering process.

The reconstruction by aWiener filter, is shown in Fig.6.4C. An other approach to solvethe problem of noise are iterativealgorithms. The idea is to guessthe ’ideal’ image, to convoluteit with the point spread functionand to compare the result withthe real image.

After an appropriate adjustmentof the guessed image, the proce-dure is started again. Fig.6.4D shows the result of 18 itera-tions of the van Cittert algorithmand Fig. 6.4 E shows the re-construction by the Richardson-Lucy maximum likelihood algo-rithm. Best results are obtainedby the Richardson-Lucy method.However none of the algorithmsare able to recover details likethe screw thread. Further im-provements in the image qualitycan be achieved by a modifica-tion of the point spread function(see ’Coded apertures in neutronradiography’ in the same report).

[1] J.C.Russ."The image processing handbook"(CRC Press, Boca Raton Ann Arbor London Tokyo,1995).


6.4 Coded Apertures in Neutron Radiography

F. Grünauer 1


Most neutron radiography fa-cilities use circular single holeapertures. Such apertures pro-vide a so called pillbox pointspread function. As the pointspread function has a crucial im-pact on the quality of recon-structed images, investigationswere carried out with regard to anoptimal aperture design. Manydifferent configurations were sim-ulated by Monte Carlo method.The test specimen is presentedin ’image deconvolution in neu-tron radiography’ in the same re-

port. A comparison of a multi-ple hole aperture (fig.6.5 E) tothe ’classic’ single hole aperture(fig. 6.5A) is shown below. Bothapertures have the same transmis-sion area. The overlap of in-formation in the projection withthe multiple hole aperture wasreduced by minimization of theautocorrelation function of thepoint spread function. The result-ing image is a superposition ofsix images (fig.6.5 G). For bothaperture configurations the imagewas reconstructed by a Wiener

filter. The reconstructed image,obtained with the multiple holeaperture (fig.6.5H), shows moredetails (e.g. the screw thread)and is less noisy than the re-constructed image, which stemsfrom the single hole arrangement(fig. 6.5 D). The quality of opti-cal systems is usually defined bythe modulation transfer function[1]. Fig. 6.5 B and Fig. 6.5 Fshow, that the modulation trans-fer function has considerably lessnear zero zones in case of themultiple hole aperture.

Figure 6.5:A) Single hole aperture, B) modulation transfer function for the single hole aperture, C) unprocessedprojection from the single hole aperture, D) reconstructed image for the single hole aperture, E) multiple holeaperture, F) modulation transfer function for the multiple hole aperture, G) unprocessed projection from themultiple hole aperture, H) reconstructed image for the multiple hole aperture

[1] Hecht E."Optik" (Addison-Wesley, Bonn München, 1989).


6.5 Non-Destructive Testing With Phase Contrast Radiography

K. Lorenz 1, E. Steichele1, E. Lehmann2, P. Vontobel2

1 Physikdepartment E21, TU München2 Paul-Scherrer-Institut, Villigen, Schweiz

Theory

The refractive index of a materialis given by the complex expres-sion n = 1− δ − iβ . The imagi-nary part of this value is respon-sible for the mass attenuation co-efficient µ of the material whichcreates absorption contrast. Thereal partδ causes a phase shift,which leads to a deviation of thewave front. If the incoming waveis approximately a plane wave(neutrons with a high transversalspatial coherence), you get in cer-tain object-detector-distances in-terference effects. Fresnel’s the-ory for diffraction, explains theincrease in contrast at edges andinterfaces in the object [1] (wherethe gradient of the phase shift per-pendicular to the neutron beam isvery high; see Fig.6.6). This ad-ditional contrast to the absorptioncontrast is called phase contrast.

Applications

Prof. Wellnitz, holder of thechair of conceptional lightweightconstruction at the FH Ingolstadt,provided us with probes of alu-minum foams. In spite of thevery small attenuation coefficient

Figure 6.6:Principle

of aluminum, it is possible tovisualize the inner structure ofthese foams with the help ofphase contrast (see Fig.6.7).

If during the casting processthe outer region of a cast compo-nent solidifies quicker than the in-ner, the shrinking during solidifi-cation can create little holes in-side the cast material. The In-stitut für Umformtechnik(UTG)lent us cast components withsuch defects inside. The sizeof those defects is often far toosmall to get enough absorptioncontrast to visualize them in aconventional radiography.

In association with Dr. Mol-lenhauer, director of the researchdepartment of the orthopedicsurgery of the Friedrich-Schiller-University Jena, we use phasecontrast neutron radiography tovisualize the interface betweenan implant and the bone. Bylooking at this interface at differ-ent times after the implantationof the prosthesis, it is possible tostudy the dependence of the heal-ing process on the surface mate-rial and the surface structure ofthe prosthesis.

Recent Topics ofResearch

A disadvantage of phase contrastradiography are the long expo-sure times compared to conven-tional radiography. This is be-cause of the need for a neutron

Figure 6.7:phase contrast radiogra-phy of an aluminum foam (madeat the PSI in Switzerland; thediameter of the visible area is42 mm)

beam with a high transversal co-herence, which is achieved by theapplication of small pinholes andlarge distances between this pin-hole and the probe. This leads toa strong reduction of the beamsintensity and thus to long expo-sure times (90min and more). Apossible solution for this problemis the utilization of coded aper-tures [2], which replaces the sin-gle pinhole.

The small diameter (<1mm) ofthe pinhole allows the applicationof neutron lenses with very smallradii. Like this, it is possibleto create lens systems with a fo-cal length in the order of meter.With such a device, the resolutionin the object plane could be im-proved by a factor of two or moreby a given detector resolution.

By making three phase con-trast radiographies in three differ-ent object-detector-distances, it


is possible to use the transportof intensity equation (TIE) tocalculate the phase shift of theneutrons on their different waysthrough the object [3]. Doing

this phase retrieval for every sin-gle projection of a tomographymakes it possible to assign everyvoxel of the tomography a phaseshift and an attenuation coeffi-

cient. Like this, it becomes pos-sible to separate materials, withsimilar attenuation coefficients.

[1] Cowley J. M.Diffraction Physics(NHPL, 1995), 3 edition.[2] Skinner G. K.Nuclear Instruments and Methods in Physics Research, 221, (1984), 33–40.[3] Gureyev T. E.Optics Communications, 147, (1998), 229–232.

6.6 NRSE coils for MIRA

N. Arend 1, R. Georgii2, H. Wagensonner2, T. Keller 3

1 TU München, Department of Physics, Institute E21, 85748 Garching2 TU München, FRM-II ZWE, 85747 Garching3 Max-Planck-Institute for Solid State Research, 70569 Stuttgart

Introduction

For the multi-MIEZE (Modula-tion of IntEnsity by Zero Effort)[1] option of MIRA, NRSE (Neu-tron Resonance Spin Echo) flip-per coils must be implemented.The standard design of thosecoils requires the neutron beamto penetrate them and thereforeinteract with the coil material.MIRA will operate in the coldneutron range, this must be takeninto account when choosing ma-terials and the design to avoidstrong absorption and scattering.

Design efforts

The NRSE flipper coils usedat other FRM-II instruments(RESEDA and TAS-NRSE) havetwo properties that make themunfavourable for MIRA: the DCcoils are made of a 0.5 mm× 8mm aluminum band which has ananodised Al2O3 coating that fea-tures a periodic structure. Thisstructure causes significant small

angle scattering at higher wave-lengths. Furthermore, since thewinding density is relatively low,high currents (≈ 100 A) must beused to achieve the nominal DCfield of 0.03 T.

In order to get around thosedrawbacks a new design isplanned for the MIRA coils. Itis based on a design proposedin [2] and developed at the MPIfür Metallforschung in Stuttgart.Thin (0.5 mm) profiles made of arigid Al alloy (see Fig. 6.8) arepiled on top each other, formingthe windings of the DC coil.

A considerably higher wind-ing density (and therefore smallercurrent) as well as improved scat-tering properties can be achieved.Several approaches are being fol-lowed concerning electrical insu-lation (e.g. sputtered glass coat-ing), contacts of the windings,and cooling. The design processis in progress. Significant im-provements seem possible com-pared to the original Stuttgart de-sign.

Figure 6.8: Simplified single DCcoil winding (cooling channelsare omitted).

B-field simulations

In order to test various designsprior to building, numerical simu-lations help finding design flawsand optimising the model. Fig.6.9 shows a B-field contour plotat the center of a DC coil of≈ 31cm width (150 windings @ 2 mmprofile thickness, 13 A current)based on the sketched profile.


Figure 6.9:B-field contour plot ofthe DC coil.The mu-metal shielding as

well as cooling and other detailsare neglected. Deep red areasmark high, blue areas lower fieldmagnitudes. This simulation wasdone using the IES© Faraday©

software package.

Science with MIRAsmulti-MIEZE option

A multi-MIEZE setup is capableof producing very short, well sep-

arated pulses from a constant, po-larised neutron beam [1]. A four-stage setup with frequency dou-bling between two subsequentlevels produces the beam struc-ture shown in Fig. 6.10 (com-pared to a single MIEZE signal).

If an additional spin flippercoil is placed directly behind thelast MIEZE level, a macroscopicspatial splitting of the spin-up (↑)and spin-down (↓) states of theneutron pulses can be achieved(Fig. 6.11) [3].

tD

I (tD)

Figure 6.10:Peaks produced by afour level multi-MIEZE (sharppeaks, compared to broad singleMIEZE signal).

tD

IF(tD)

⟨↑⟩⟨↓⟩

13 µs

Figure 6.11:Spatial splitting of thespin-up (↑) and spin-down (↓)states.

This "longitudinal Stern-Gerlacheffect" demonstrates the predic-tions of quantum mechanics incontrast to classical physics.

Besides the demonstration ofthis basic quantum-mechanicaleffect, the so prepared beam mayserve as an experimental basis forstudying interference and coher-ence phenomena using cold neu-trons.

[1] Keller T., Golub R., Gähler R. In: Pike R., Sabatier P. (Editors),Scattering and Inverse Scatteringin Pure and Applied Science, 1264–86 (Academic Press, San Diego, CA, 2002).

[2] Wahl M. Aufbau einer Messanordnung zur Neutronenreflektometrie. Master’s thesis, Max-Planck-Institut für Metallforschung, Stuttgart, Germany/Institut für Theoretische und Angewandte Physik,Universität Stuttgart, Germany (2003).

[3] Arend N., Gähler R., Keller T., Georgii R., Hils T., Böni P. Classical and quantum-mechanicalpicture of NRSE - Measuring the longitudinal Stern-Gerlach effect by means of TOF methods(2003). Submitted to Physics Letters A.


6.7 Steps Towards Dynamic Neutron Radiography of aCombustion Engine

J. Brunner 1, G. Frei 2, R. Gähler4, A. Gildemeister3, E. Lehmann2, A. Hillenbach 3, B. Schillinger1

1 Physik Department E21, TU München2 Paul Scherrer Institut, 5232 Villigen, Switzerland3 Universität Heidelberg, Physikalisches Institut, 69120 Heidelberg, Germany4 Institute Laue Langevin, 38042 Grenoble, France

The Dynamic NR is a newtechnique, which is applied suc-cessfully for non-destructive test-ing for time dependent phenom-ena with a similar resolution anda similar contrast as in conven-tional NR. Of special interest isthe investigation of the injectionprocess in a running combustionengine under real conditions.

For obtaining a 16-bit conven-tional radiography image at a typ-ical flux of 106 n/cm2s one hasto integrate scintillator light onthe CCD chip typically for someseconds. Of course the samplethickness, the sample material aswell as the detector efficiency de-termine the exact time. Addi-tionally the readout time, depend-ing on the dynamic range and thepixel number of the CCD, puts alower limit to the highest achiev-able frame rate.

The observation of objects inthe millisecond time scale is im-possible with standard methods.A possible solution for repetitiveprocesses can be stroboscopicimaging. That means a synchro-nization of the detector with therepetitive process and an integra-tion of 1000 images taken at thesame millisecond time windowof the cycle. That way a virtualopening time of some secondscan be realized as long as thenoise is constant within time (i.e.

negligible readout noise) and notdetermined by the system.

With this method two combus-tion engines of aluminum towedby a electric motor were ob-served at ILL. The achieved spa-tial resolution was 1 mm becauseof the beam unsharpness.

Figure 6.12:Two images of the run-ning cycle of a motorbike at 800rpm, 250ms with 60 image accu-mulations

Figure 6.13:Piston cooling of a fourstroke car combustion engine canbe observed well by neutron ra-diography, t=200ms with 120 ac-cumulations, image size is 24cmx 24cmThis dynamic neutron investi-

gations were performed with andetector system from PSI at avery high flux test beam line atILL. The PSI detector consistedof a peltier cooled CCD camerausing a multi channel plate as animage intensifier coupled on theCCD-chip with fiber optics.

With this equipment and thehigh flux thermal neutron beamof 109 n/cm2s at ILL extremelyhigh detection sensitivity waspossible. The necessary fast trig-gering is done by the image inten-sifier. Since the scintillator decaytime to 10% is 85µs, that is theorder of magnitude for the limitof the time resolution.

In the false color image Fig.6.14 the small attenuation of theinjection liquid cloud is clearlyvisible. The next step will besuch an observation in a runningengine under real conditions.

The obtained results should bea starting point to fit the differ-ent requirements of car producersin respect to fuel injection, lubri-cant distribution, mechanical sta-bility and operation control. Sim-ilar inspections will be possiblefor all devices with repetitive op-eration principle. At the FRMII the ANTARES tomography sta-tion with a high neutron flux of108 n/cm2s and a high spatial res-olution will allow such measure-ments soon. [1, 2]


Figure 6.14:Common Rail Diesel injection nozzle (500 bar) in action, 5000neutron radiography frames of 100ms added by software, delay to thetrigger, 600µs, 700µs, 800µs, image size is 6.2cm x 8.2cm

[1] Schilllinger B. Neue Entwicklungen zu Radiographie and Tomographie mit thermischen Neutro-nen. Ph.D. thesis, Technischen Universität München (200?).

[2] Balasko M.Physica B: Condensed Matter, 234-236, (1997), 1033–1034.

6.8 Development of a Fuel Element with Reduced Enrichment forthe FRM-II

K. Böning 1


On May 2, 2003, the Bavar-ian State Ministry of Environ-ment (StMLU) - on instructionsreceived from the Federal Min-istry of Environment (BMU) -awarded the FRM-II its 3rd par-tial nuclear license permittingnuclear startup and full powerroutine operation of the reac-tor. However, this long awaitedand most welcome final licensealso involved the condition thatthe FRM-II should develop, un-til the end of the year 2010, anew fuel element allowing theFRM-II to continue its operationwith medium enriched uranium(MEU, with a maximum of 50% U235) instead of the presenthighly enriched uranium (HEU,with 93 % U235). This means

that - as a first step towards thisgoal - an advanced fuel with amuch higher density of uraniumthan presently available must bedeveloped so that the reductionof enrichment could be compen-sated by an increase in density.Only in this way the dimensionsof the fuel element can be keptunchanged, the replacement ofthe central part of the reactorfacility can be avoided and thecondition of "any penalty to theFRM-II being marginal only" canbe obeyed.

After the issue of the 3rd nu-clear permit an agreement be-tween the Bavarian State Min-istry of Science (StMWFK) andthe Federal Ministry of Science(BMBF), which has been negoti-

ated already two years ago andwhich also settled the budgetrequirements, has been signedon May 30, 2003. In con-sequence an international work-ing group has been establishedat the Technische UniversitätMünchen (TUM) comprising rep-resentatives of the FRM-II, ofthe fuel element manufacturerCERCA (France) and of theresearch reactor planning com-pany Siemens/Framatome-ANP.The first meeting of this "MEUWorking Group" was performedin Garching on July 29, 2003,where the break down of theproject into three phases of 2 ½years each has been agreed upon(see table6.1).

Being well aware that this


schedule is extremely tough theworking group will do everythingto render this project a success.In September 2003 the BMBF is-sued the official notification ofhaving allocated its share of thebudget for the project phase I.

The second meeting of theMEU working group was held inGarching on October 17, 2003.The FRM-II reported on neutron-ics calculations which showedthat an uranium density in theMEU fuel of 8.0 - 8.5 g/cm3

will be required for a fuel ele-ment with 50 % enrichment (thepresent density for HEU is 3.0g/cm3); the maximum obtainedfission density in the fuel meatwill be of the order of 2.3×1021

fissions/cm3. It was decided toask CERCA to submit an offerconcerning the supply of MEUand the fabrication of 4 test irra-diation plates each with 8.0 den-sity but different metallurgy andof 2 spare plates with 7.0 g/cm3

Phase I: 07.2003 - 12.2005: Fuel development and basicquestions

Phase II: 01.2006 - 06.2008: Fuel specification and design offuel element

Phase III: 07.2008 - 12.2010: Fuel element fabrication and li-censing

Table 6.1:Time table for the MEU fuel element development

for the case that some of the testirradiations should fail. The testirradiations will be performed atthe OSIRIS materials testing re-actor in Saclay which is operatedby the French "Commissariat al’Energie Atomique" (CEA) andwhich has already been used forthe irradiation of HEU fuel platesof the FRM-II some years ago.On November 24, 2003, TUMand CERCA met with CEA atSaclay to negotiate the details ofa contract concerning the irradia-tion of a total of 4 test plates (2 ofeach density) in the IRIS facilityof OSIRIS, post irradiation exam-inations (PIEs) of probably two

plates in a hot cell at Cadarache,and the final disposal of theplates which have become highlyradioactive under irradiation.The final offers and updated draftcontracts of CERCA and CEAare expected shortly. In ordernot to loose time and anticipatingthese contracts we have sent, onDecember 12, 2003, a letter ofconfirmation calling on CERCAto begin immediately with theproduction of the 50 % enrichedMEU (as to be obtained frommixing 93 % and 20 % enricheduranium) and with the first stepsto fabricate the irradiation testplates.

80 7 Scientific Highlights

7 Scientific Highlights

7.1 Pressure Induced Magnetic and Valence Transitions inYbMn2Ge2

M. Hofmann 1, S.J. Campbell2,1, P. Link 1

1 ZWE FRM-II, Technische Universität München, 85748 Garching2 School of Physical, Environmental and Mathematical Sciences, The University of New South Wales,

ADFA, Canberra, ACT 2600, Australia

Introduction

Rare-earth intermetallic com-pounds containing ytterbium ex-hibit a wide range of interestingand unusual physical and mag-netic properties. This occursmainly as a result of their mixedvalence states (II/III) or changesfrom one valence state to theother. Applied pressure is likelyto cause changes in the valencecharacter of systems exhibitingintermediate valence and is there-fore correspondingly expected toinfluence the magnetic behaviour[1]. This was highlighted in re-cent pressure magnetisation mea-surements [2] which revealedanomalous changes in the mag-netic behaviour of YbMn2Ge2

around a critical pressure ofabout 1.25 GPa. At ambient pres-sure the Yb mean valence wasestimated to be≈ 2.35 in thiscompound [3, 4] and therefore itis expected that only the Mn sub-lattice will order magnetically.

We have previously demon-strated that YbMn2Ge2 crystal-lizes in the tetragonal ThCr2Si2type structure; it has a pla-nar antiferromagnetic structure

below TN1 ≈ 510K, transform-ing to a canted antiferromagneticstructure below TN2 ≈ 185K[5]. Our recent neutron andx-ray diffraction experiments re-veal that YbMn2Ge2 under pres-sure exhibits distinct changes inthe magnetic structure along witha sharp decrease in the a-latticeparameter above 1.5 GPa [6]. Inaddition we also observe

0 1 2 3 41 6 0

1 6 4

1 6 8

1 7 2

1 7 6

1 8 0

��

��

��

��

�

��

��

��

Figure 7.1:The variation of the unit cell volume of YbMn2Ge2 with pres-sure as determined by the x-ray diffraction measurements at ambient tem-perature (the lines are fits to the data as discussed in the text).

reversible peak splitting charac-teristic for a valence transition offirst order around the critical pres-sure range p≈ 1.25 GPa.

Valence Transition

Figure 7.1 shows the variationof the unit cell volume ofYbMn2Ge2 as a function of pres-sure as determined from our

7 Scientific Highlights 81

x-ray diffraction investigation atambient temperature [6]. The keyfeature is the step like behaviourof the volume at around p≈ 1–1.5 GPa. This is caused primar-ily by a significant decrease in thea-lattice parameter (≈ 3%) withthe c-lattice parameter found tovary only slightly over the entirepressure range. Given that thedimensions of the basal plane inthis tetragonal ThCr2Si2 - typestructure are determined primar-ily by the radii of the rare earthatoms and that, in general, ap-plied pressure favours change inthe valence character of com-pounds containing Ce, Eu andYb e.g. [1], this pronounceddecrease in the lattice parameterprovides evidence for a pressure-induced valence transition from adivalent-like Yb ionic state to aYb3+ state in YbMn2Ge2. Thisbehaviour fits in well with thesignificant changes in magneticproperties evident at pressure p≈ 1.25 GPa in the magnetisationmeasurements of Fujiwara et al[2].

In order to derive the bulk mod-ulus using Murnaghan’s equationof state [7], because of the strongvariation in the volume, we distin-guish between lower and higherpressure regions separated at≈ 1–1.5 GPa. As shown in Figure7.1, between 0 GPa and 1 GPaYbMn2Ge2 has a much smallervalue for B0 = 37(6) GPa withV0 = 180.0(4) Å3, than at pres-sures above p≈ 1.5 GPa wherethe fit yields B0 = 85(6) andV0 = 169.4(3) Å3. This lat-ter value of V0 is close to the

6 0 8 0 1 0 0 1 2 0

0

2 5 0

5 0 0

��

��

��

��

��

��

��

��

�

Inten

sity [a

.u.]

�θ��

��

Figure 7.2:Neutron diffraction patterns of YbMn2Ge2 at p≈ 1.35 GPa be-tween T = 1.5 - 280 K. Indicated are the (110) reflections with latticespacings a(2+)≈ 4.03 Å and a(3+)≈ 3.96 Å as discussed in the text.

value V = 167.9 Å3 [4] calcu-lated for a hypothetical trivalentYbMn2Ge2 state, thus indicatinga valence change to almost triva-lent Yb between p≈ 1 GPaand 1.5 GPa. This is consis-tent with recent LIII -edge mea-surements on YbMn2Ge2 whichindicate a first-order transition p≈ 1.25 GPa at which pressurethe Yb mean valence changesabruptly from 2.4 to a mean va-lence of about 2.85 at p = 1.3 GPa[8].

Further evidence for this va-lence transition is found in theanalysis of our neutron diffrac-tion data which were obtained atp ≈ 1.35 GPa (figure7.2). Theexperiments were carried out onthe G6.1 diffractometer at LLB,Saclay. As shown in figure7.2, atT = 280 K we observe peak split-ting of the main structural peaks;

in particular we observe two(110) reflections which indicatethe presence of two phases withdifferent a-lattice constants, e.g.a(2+)≈ 4.03 Å and a(3+)≈ 3.96Å. As stated above a change inthe rare earth ionic radii will af-fect mainly the basal plane, thelarger lattice spacing a(2+)≈4.03 Å, therefore corresponds toa valence state of divalent charac-ter while the smaller lattice spac-ing, a(3+)≈ 3.96 Å, is consis-tent with trivalent Yb. Basedon the relative peak intensities,the fractions of the Yb2+ likeand the Yb3+ phases are esti-mated as 35 % and 65 % respec-tively. On lowering the temper-ature the phase fraction of triva-lent YbMn2Ge2 decreases to 40% with the change found to be re-versible with temperature.


Magnetic structure athigh pressure

The neutron diffraction pattern ofYbMn2Ge2 at ambient pressure(figure 7.3) shows the charac-teristic crystallographic featuresof the tetragonal ThCr2Si2 struc-ture. As well, the magnetic reflec-tion (101) is clearly discernible,as expected for the planar an-tiferromagnetic arrangement ofthe Mn moments at ambient

01 x 1 0 4

2 x 1 0 4

3 x 1 0 4

4 x 1 0 4

5 x 1 0 4

6 x 1 0 4

5 0 7 5 1 0 0 1 2 5

0

5 0 0

1 0 0 0

��

��

��

��

�

Norm

alised

coun

ts

��

Norm

alised

coun

ts

�θ��

Figure 7.3:Neutron diffraction patterns of YbMn2Ge2 at: (upper) ambi-ent pressure and temperature; (lower) 2.7 GPa; T = 280 K. The linesare Rietveld refinement fits to the data with the markers indicating thepeak position of the nuclear (bottom) and magnetic (top) reflections ofYbMn2Ge2. The small peak at≈ 82 in the pattern at 2.7 GPa resultsfrom an instrumental effect.

temperature [3]. The refined mag-netic moment at T = 290 K is 2.5µB, increasing to 3.4µB at T =1.5 K consistent with our earlierfindings [3]. The lattice parame-ters of YbMn2Ge2 at T = 290 Kare a = 4.083 Å and c = 10.902 Å[9].

By comparison, at an appliedpressure p≈ 2.7 GPa the (101)reflection - the main magnetic re-flection in the planar antiferro-magnetism of YbMn2Ge2 - is

excluded from the diffraction pat-tern and disappears almost com-pletely. As well, the appearanceof an additional magnetic reflec-tion at 2θ ≈ 123 indicates a newmagnetic ordering scheme. Thisreflection can be indexed as (111)on the basis of a primitive tetrag-onal cell. The resulting struc-ture closely resembles the mag-netic structure of YbMn2Si2 (oftrivalent Yb) at ambient condi-tions. Indeed YbMn2Ge2 at apressure of 2.7 GPa is found toexhibit similar lattice constantsand, more importantly, similarMn-Mn bond distances to thoseof trivalent YbMn2Si2 at roomtemperature (dMn−Mn = 2.768 Åin YbMn2Ge2 and dMn−Mn =2.757 Å in YbMn2Si2). In this lat-ter structure, ferromagneticallycoupled Mn layers are coupledantiferromagnetically along the c-axis [10] with the moment direc-tion along the z direction. Analy-sis of the neutron diffraction pat-tern of YbMn2Ge2 (at T = 280 Kand at a pressure of p = 2.7 GPa)in terms of the collinear antifer-romagnetic structure of trivalentYbMn2Si2 yields a moment valueof µ ≈ 1.3 µB with refined latticeconstants of a = 3.940 Å and c =10.786 Å.

The current investigationsdemonstrate clearly the extrainsight which variable temper-ature neutron diffraction mea-surements carried out under pres-sure can provide in the study ofcompounds exhibiting valencechanges. Our continuing studiesof a series of Yb and Eu basedcompounds indicate clearly thepotential and prospects for incor-porating pressure-cell measure-ments into the routine investiga-


tions available to users at FRM-II.

SJC acknowledges the sup-

port of the Alexander von Hum-boldt Foundation during renewalof his Research Fellowship at

FRM-II, Technische UniversitätMünchen.

[1] Thompson J., Lawrence J. In: Schneider K., Eyring L., Lander H., Choppin G. (Editors),Handbookon the Physics and Chemistry of Rare Earths, volume 19 (Elsevier, 1994).

[2] Fujiwara T., Fujii H., Uwatoko Y., Koyami K., Motokawa M., Shigeoka T.Acta Phys. Pol., 34,(2003), 1541.

[3] Hofmann M., Campbell S. J., Edge A. V. J.Appl. Phys., 74, (2002), S713.[4] Szytula A., Penc B., Jezierski A., Hofmann M., Campbell S. J.Acta Phys. Pol., 34, (2003), 1561.[5] Hofmann M., Campbell S. J., Szytula A.J. Alloys Comp., 311, (2000), 137.[6] Hofmann M., Campbell S. J., Link P., Goncharenko I., Knorr K.Physica B.[7] Murnaghan F. D.Finite Deformation of an Elastic Solid(Dover, New York, 1967).[8] Fujiwara T., Uwatoko Y., Fujii H., Koyami K., Motokawa M., Shigeoka T.J. Magn. Magn. Mater.

In press.[9] Hofmann M., Goncharenko I., Link P., Campbell S. J. (in preparation 2004).

[10] Hofmann M., Campbell S. J., Edge A. V. J., Studer A. J.J. Phys.: Condens. Matter, 13, (2001),9773.

7.2 Critical Exponents in MnSi

R. Georgii1, P. Böni2, D. Lamago1, S. Stüber1, S. V.Grigoriev3, A.I. Okorokov 3, H. Eckerlebe4,P.K. Pranzas4, B. Roessli5, W.E. Fischer5

1 ZWE FRM-II, TU München2 E21, Physik-Department, TU München3 Petersburg Nuclear Physics Institute, Gatchina, St.Petersburg 188300, Russia4 GKSS Forschungszentrum, 21502 Geesthacht5 Laboratory for Neutron Scattering, ETH Zürich & Paul Scherrer Institute, CH-5232 Villigen PSI,

Switzerland

The chiral fluctuations in theweak itinerant magnet MnSiwere studied by means of smallangle neutron scattering with po-larized neutrons (see JB 2002for details). Due to the lackof a centre of symmetry themagnetic moments are arrangedalong a left-handed spiral due tothe Dzyaloshinskii-Moriya (DM)interaction. The experimentsshow that the incommensuratemagnetic peaks evolve with in-creasing temperature into diffusescattering that is mainly concen-trated in a ring with the radius

qc = 0.039 Å−1 centred at the po-sition of the direct beamq = 0.

For the determination of thecritical exponents we started tointerpret the data by extractingdifference spectra of the twobeam polarisations. Therefore,the effect of background scatter-ing is eliminated and only the fullanti-symmetric part of the mag-netic cross section is considered.The spectra are integrated alonga ring of constantq for both po-larisations independently (i.e. theintegration is performed along ahalf circle of 1800) and the in-

tensity is plotted againstq. Theresulting structure is fitted witha Lorentzian peak. The two pa-rameters of this fit, the width andthe intensity, are plotted againstthe temperature as shown in Fig-ure 7.4. From these two plotsthe following critical exponentscan now be extracted usingTc =(28.7±0.05) K for all fits:

As the imaginary part ofthe paramagnetic susceptibility isproportional to the neutron mag-netic scattering function we canderive the critical parameterγ forT ≥ Tc directly from the intensity


Figure 7.4:The intensity (a) and width (b) of the Lorentzian fits versus thetemperature. From the intensity plot below Tc the critical exponentγ forthe magnetisation M and above Tc the critical exponentβ for the param-agnetic susceptibilityχ is obtained. From the width one gets the criticalexponentν for κ, the inverse of the correlation length.

plot. For MnSi we obtain a valueγ = 0.61±0.01.

As the intensity of the scatter-ing in the ordered phase is propor-tional to the square of the mag-netisation we get the critical pa-rameter 2β for T ≤Tc from the in-tensity plot. For MnSi we obtainβ = 0.218± 0.016. The widthof the peak is proportional to theinverse of the correlation length,therefore we obtainν from thewidth of the peak. We findν =0.541±0.011.

These results are so far onlypreliminary as we intend to re-measure MnSi for obtaining bet-ter statistics and also need toinclude the effects of the finiteinstrument resolution. But afew conclusions can already bedrawn: While the values forβand ν are similar to the valuesof other magnetic systems [1] thevalue forγ clearly is smaller. An-other interesting observation isthat ford = 2 we get forγ +2β −d · ν = −0.036± 0.017, indicat-ing a two-dimensional magneticsystem.

[1] B.D. Gaulin M. C., Mason T.Physica B, 156&157, (1989), 244.


7.3 Dipolar Anisotropy of Spin Waves in Heisenberg FerromagnetEuS Investigated by Small-angle Polarized Neutron Scattering

D. Lamago1,4, O. Okorokov 2, S. Grigoriev2, P. Böni1, R. Georgii1, H. Eckerlebe3, P. K. Pranzas3

1 Physikdepartment E21, TU München2 Petersburg Nuclear Physics Institute, Gatchina, 188350, Russia3 GKSS-Forschungszentrum, Max-Planck-Str.,21502 Geesthacht, Germany4 ZWE FRM-II, TU München

Introduction

Spin dynamics have been stud-ied in the isotropic ferromagnetEuS by means of small-angle po-larized neutrons scattering in thetemperature range 14K ≤ TC =16.6 K ≤ 35K in external mag-netic fields of H=0-250 mT in-clined about ϕ = 45◦ relativeto the incident beam. Isotropicferromagnets have been subjectof extensive studies. Due toits strongly localized momentsand low anisotropy fields, EuSis a very good realization of aHeisenberg ferromagnet. In ad-dition EuS is of particular inter-est because of the large momentof the Eu2+ ions of 7µB lead-ing to the strong dipolar inter-actions, which are not negligi-ble in comparison with the ex-change interaction.The goal ofthis research has been to investi-gate the influence of the dipole-dipole interaction on the criticalbehaviour of the system at tem-peratures above the Curie tem-perature (TC = 16.5 K).The tech-nique apply for this experimentmake use of the asymmetry ofthe small-angle scattering of po-larized neutrons[1, 2] .

Theoretical Foundation

According to [2] one can observethe left-right asymmetry of polar-ized neutron scattering by mea-suring the difference in the inten-sities of the neutrons which arescattered through a small angleθ and which are originally po-larized along the magnetizationof the sample and opposite to it:∆ I(θ) = I(θ ,P0) − I(θ ,−P0),where P0 is the initial polar-ization. This quantity canbe written in the form [3]:

(7.1) ∆I(θ)≈∫

A2m(em)(eP0)Sa

q,ω +AmαM(P0m⊥)φq,ωdω

The first term describes themagnetic scattering while the sec-ond describes an interference ofmagnetic and nuclear scattering.Am = γe2/meC2 is the amplitudeof the magnetic scattering of theatom. α is the expectation valueof the nuclear amplitude,e =q/q,m= M/M, M is the magne-tization of the sample andm⊥ =m− e(em). The function Sa

q,ω

describes the antisymmetric spincorrelations andφq,ω is the dy-namic nuclear structure factor.

In the case of small-angle scat-tering where ω is very smallcompared to the energy of inci-

dent neutrons, the scattering byspin waves can be separated ex-perimentally from the contribu-tion of magnetic nuclear scat-tering. For this purpose oneshould perform the measurementof ∆I(±θ ,P0) with the polariza-tion directed along and oppositeto the magnetization of the sam-ple and form the difference∆IA =12[∆I(θ ,P0)−∆I(−θ ,P0)].

Experiments And Results

The SAPNS experiments were

performed using the SANS-2 ofthe research reactor FRG-1 atthe GKSS-Forschungszentrum inGeesthacht(Germany). All mea-surements were conducted us-ing polarized neutrons (P0 =0.95) of fixed wavelengthλ =5.8 (∆λ/λ = 0.1) with angulardivergence of 10 mrad. The scat-tered neutrons were detected by aposition sensitive detector 258×258 mm2. The isotopically en-riched sample153EuS was com-posed of roughly 100 single crys-tals, aligned such that the over-all mosaic wasη ≈ 0.75◦. Thesample was mounted inside a cry-


omagnet inclined by an angleϕ = 45◦ relative to the incidentbeam and providing a field upto 250 mT. The scattering inten-sity were measured in the tem-perature range fromT = 14 Kto T = 50 K. With the de-scribed experimental setup onecan observe the left-right asym-metry of neutron scattering andthus investigate the critical dy-namics by triple-spin correlationsin the magnetic field and measurewith high accuracy the parame-ters of spin waves in ferromag-netic phase [4, 5].

The two-spin correlation

In the measurements shown inFig. 7.5 and 7.6 one can no-tice the changes in both the quasi-elastic and the inelastic scatteringcontributions due to the criticalspin fluctuations and to the spin-wave excitations respectively.

The measured intensity of thescattered neutrons is well de-scribed by the Ornstein-Zernickeexpression:

(7.2) I(q) = A(q2 +κ2)−1,

where A is the scattering ampli-tude andκ is the inverse corre-lation length. Fig.7.5 showsthe temperature dependence ofthe correlation lengthRC = 1

κ.

For temperatures aboveTC andup to the highest temperaturesof our experiments, the lineshapes are well described bythe spin correlation theories.RC

increases while temperature ap-proaches the Curie temperatureTC = 16.5 K from above and de-creases with further lowering ofthe temperature. The correlationlength decreases with increas-ing magnetic field as shown inFig.7.6. The behaviour ofRC canbe explained through the strongfield regime, which is determinedaccording to Ref.[6] by the con-dition TC(κCa0)z = gµH. Here zis the critical field exponent anda0 is a constant of order of 1 Å.From the field dependence of thecorrelation length atTC the valueof the parameter1z = 0.65±0.02was evaluated. The devia-tion from the theoretical value(1/z = 0.4) of a pure Heisen-berg ferromagnet is attributed

to the dipolar interactions.

Figure 7.5: Influence of tempera-ture on the correlation length

Figure 7.6: Magnetic field depen-dence of the correlation radius atthe Curie temperature

[1] Okorokov A. I. Physika B, 204-205, (2000), 276–278.[2] Okorokov A. JETP Lett., 48, (1986), 8.[3] Gukasov A.Physika B, 267-268, (1999), 97–105.[4] Toperverg B., Deriglazov V., Mikhailova V.Physika B, 183, (1993), 326.[5] Deriglazov V.Physika B, 180-181, (1992), 262–264.[6] Maleyev S.Physics Review, 8, (1987), 323.


7.4 Submicrometer X-ray Tomography with Doubly AsymmetricalBragg Diffraction at SLS

M. Stampanoni1, G. L. Borchert 2, P. Rüegsegger3, R. Abela1, J. Rebelo-Kornmeier2

1 SLS PSI, Villigen, Schweiz2 ZWE FRM-II, TU München3 IBT ETH, Zürich, Schweiz

Introduction

Microtomography is a challeng-ing research domain being ap-proached with neutron as wellas X-ray techniques. By exploit-ing the high intensity of third-generation synchrotron sourcesand combining hard X-ray mi-croscopy with tomographic tech-niques, it is possible to obtainvolumetric information of a sam-ple in the micrometer or evensub-micrometer scale with min-imal sample preparation. Thechallenge to trespass the thresh-old of one micrometer spatialresolution with hard X-rays hasspurred the development of de-tectors optimized for spatial res-olutions to the detriment of ef-ficiency. In fact, state-of-the-

Figure 7.7: X-ray optics ofthe Bragg magnifier showing themain crystals and the objects andimage planes. ”O” describes theincident beam coordinates, ”H”the diffracted one. Inset: copla-nar, asymmetric one-dimensionalBragg diffraction; m is the magni-fication factor.

art devices, which convert X-raysto visible light with a thin scin-tillator and project them ontothe CCD with a microscope op-tics, claim a spatial resolutionaround 1 micrometer with effi-ciency of less than 1% at 30keV [1]. This, of course, im-plies very large exposure time.As a consequence, the scien-tific development aims to per-form tomographic scans (usually500-1000 two-dimensional radio-graphic projections have to be ac-quired) with sub-micrometer res-olution within minutes. At theMaterials Science Beamline 4Sof the SLS we installed a noveldetector, called Bragg Magnifierand described in detail in Ref.[2], which efficiently acquiresdistortion-free, magnified hard X-ray projection images of a samplewith spatial resolutions around250 nm.

Working principle

The Bragg Magnifier performstwo-dimensional magnificationin the X-ray regime exploitingthe well know principle of asym-metrical Bragg diffraction fromtwo crossed flat crystals, seeinset of Fig. 7.7. The inten-sity distribution of the beamcontaining absorption informa-tion about the sample is col-lected by the optics and enlarged

Figure 7.8: X-ray optics of theBragg Magnifier. Both crystals inKirkpatrik-Baez geometry as wellas the entrance of the 1:1 tandemoptics are clearly visible.

according to X-ray dynamicaldiffraction theory. Let us definethe magnification factor as m =sin(θB + α)/sin(θB - α) whereθB is the Bragg angle andα isthe asymmetry angle, i.e. theangle between the crystal sur-face and its lattice planes. Itcan be shown [3] that the angu-lar divergence of the incident (ωO ) and diffracted (ωH ) beamare related to each other byωH

= 1/|m| ·ωO For SO and SH be-ing the spatial cross-sectionsof the incident and diffractedbeam, Liouville’s theorem re-quires that SO ·ωO = SH ·ωH andhenceSH = |m| · SO.Therefore if|m| > 1 the range of total reflec-tion for the emergent beam is1/|m| times smaller than that ofthe incident beam, while its spa-tial cross section is|m| timesgreater (Fankuchen effect). Thisallows enlarging the radiographicprojection of the sample.


Realization and firstexperiments

Si(220) crystals with an asymme-try angle of α = 8◦ have beenchosen as X-ray optics. They arefixed on high-precision goniome-ter heads which allow pitchingand rolling within an angular re-producibility of 2” as well asrocking with an angular reso-lution of 0.05”. The magni-fied X-ray image is collected byan high-efficiency position sen-sitive detector and converted tovisible light by a 35x35 mm2CsI(Tl) scintillator of 300 mi-crometer thickness. The gen-erated light is projected by anhigh-efficiency 1:1 tandem opticsonto a high performance CCD.The new CCD camera featuresa 2048x2048 full frame AtmelTHX7899M chip with a full wellcapacity of more than 250’000electrons. Pixel frequency is 3.3MHz, resulting in a readout timeof 320 ms for a 2k x 2k im-age. The CCD-chip is Peltier andwater cooled to -24◦C , dark-current is 6 e−/pix/s at -22◦C and

Figure 7.9:High-resolution radiogra-phy of a human bone trabecula.Magnification is 50x50 and thefield of view is 575 x 575µm2.

the RMS noise in dark is 29 e−

The effective dynamic range is>5000:1. Sub-area readout or on-chip binning allows frame ratesup to 4-5 Hz for exposure time of50 ms. Fig.7.8shows the BraggMagnifier installed at the micro-tomography station of the Materi-als Science Beamline 4S.

Tuning the X-ray energy from21.1 keV up to 22.75 keVand rocking both crystals accord-ingly, we obtained magnificationfactors ranging from 20x20 to100x100. By setting the energyto 22.34 keV, the Bragg Mag-nifier produces a magnificationof approximately 50x50 result-ing in a theoretical pixel size of280x280 nm2. With these pa-rameters we acquired a radiogra-phy of a human bone trabecula,which is shown in Fig.7.9.

In the inset of Fig.7.9, Fresnelfringes can clearly be observed.These are principally due to thecoherent nature of the beam. Asa consequence, and according totheoretical predictions [4], the im-age has to be interpreted as anedge-enhanced projection.

From the field of materialsscience we studied a sample ofcarbon fibre reinforced materialto be used in space craft ap-plication. With the same set-tings of the parameters we ac-quired 500 two-dimensional pro-jections of the sample in equidis-tant angular intervals from 0◦to180◦. The exposure time was5 seconds for each projectionresulting in a total scan timeof approximately 40 min. Flatand dark field corrected datahave been reconstructed with astandard filtered-backprojection

Figure 7.10:A three-dimensional re-construction of the carbon fibre re-inforced sample. The fibres witha diameter of 4µm are clearly re-solved.

algorithm [5]. The result-ing three-dimensional image isshown in Fig.7.10.

Conclusion and outlook

The reliability of double asym-metrical Bragg diffraction fromflat crystals has been demon-strated and the Bragg Magnifierproduces very good X-ray im-ages. Its basic principle is appli-cable at lower energies, and thiseven with fewer difficulties sincethe crystals acceptance is higher,increasing the flexibility of theinstrumentation. With the con-crete perspective to reach voxel’ssize around 100 nm3, completelyunexplored dimensions will beaccessible to computer tomogra-phy.


[1] Koch A., et al.J. Opt. Soc. Am. A, 15.[2] Stampanoni M., et al.J. Appl. Phys., 92, (2002), 7630–7635.[3] Caciuffo R., et al.Physics Reports, 152.[4] Spal R.Phys. Rev. Lett, 86, (2002), 3044.[5] Kak A. X., Slaney.Principles of Computerized Tomographic Imaging(IEEE, New York, 1988).

7.5 Short-time Stroboscopic Neutron Imaging on a RotatingEngine

B. Schillinger1,2, H. Abele4, J. Brunner 1,2, G. Frei 3, R. Gähler 5, A. Gildemeister4, A. Hillenbach 4,E. Lehmann3, P. Vontobel3

1 ZWE FRM-II, TU München2 Phsikdepartment E21, TU München3 Paul-Scherrer-Institut, Villigen PSI, Switzerland4 Universität Heidelberg, Faculty for Physics, Germany5 Institut Laue-Langevin, Grenoble, France

Abstract

Todays neutron sources do notdeliver sufficient flux to exam-ine singular short-time events inthe millisecond range by neu-tron radiography. However, peri-odic processes can be examinedif a triggered accumulating detec-tor collects information of identi-cal time windows and positionsover several cycles of the process.At the intense neutron beam H9of ILL Grenoble, an electricallydriven BMW engine was exam-ined at 1000 rpm with time res-olution of 200 microseconds.

Detection systems

So far, the only feasible detectionsystems found are interline frametransfer CCDs or CCD cameraswith gated image intensifiers, incombination with a neutron sen-sitive scintillation screen and mir-ror.

For interline frame transfer

CCD cameras, the CCD chip areaconsists of alternate light sensi-tive pixel columns and maskedvertical shift registers. Image in-formation from the light sensitivepixels can be transferred into themasked, light-insensitive shiftregister with a single clock cy-cle, enabling for exposure timesin the range of ten microseconds.This process can be repeated sev-eral times (with flushing the im-age pixels between exposures) be-fore the whole image informationis read out via the shift registers.The advantage of this method isthat no additional noise is intro-duced by an image intensifier, thedisadvantage is that the numberof on-chip accumulations is lim-ited. Additional frame accumu-lations can be done off-line inthe computer, but their statisticsis limited by the digitization pro-cess of the individual frames.

With gated image intensifiers,theoretical gating times of downto 3 ns can be achieved, how-

ever the decay time to 10% ofthe employed scintillator is 85 mi-croseconds, limiting the achiev-able time resolution. The advan-tage of intensified CCDs is theirhigher pixel capacity which en-ables for the accumulation of sev-eral hundred gating periods be-fore readout. They are limitedby the additional quantum noisegenerated by the electron multi-plication in the intensifier. Up tonow, there is no clear superiorityof one system visible.

For both systems, less than onepercent of the total neutron flu-ence is used for imaging, result-ing in very high activation of thesample. A different approach isbeing tried with a chopper wheelin Japan, which drastically re-duces activation, but is rather in-flexible in adjusting the time win-dow.


Measurements

The beam line H9 behind theLohengrin experiment at ILLGrenoble is the most intenseneutron radiography beam inthe world, with a flux of 2∗109 n/cm2s and a collimationof L/D=110. ILL, Univer-sität Heidelberg, Paul-Scherrer-Institut and TU München collab-orated on the measurement ofan electrically driven four-pistonBMW engine. The engine wasdriven by a 2 kW electro mo-tor, mounted on a vertical trans-lation stage. Since water coolingwas not possible, the spark plugswere removed to reduce drag andheat production.

Figure 7.11:Single frame of a sequence of images of a running car engine,rotated at 1000 rpm. All components are clearly visible (valves, cylinderhead, connecting rod, piston, piston pin, piston ring). Of highest interestin this run was the oil injection from below for cooling the piston.

The detection system was aMCP intensified CCD camera PI-max with a Peltier cooled chip(1300 * 1024 pixels) with 16 bitdigitization. The usable dynamicrange is limited by the inherentnoise of the intensifier.

The full cycle of this four-stroke engine running at 1000rpm was split into 120 individualframes over 2 rotations, 150 indi-vidual images were recorded asan on-chip accumulation of a 200microseconds exposure each. Inthis way, the total exposure timefor the full run was in the orderof 18 minutes only. The field ofview was 24 cm * 24 cm. The ob-servation area could be varied bydisplacement of the full set-up.

Figure 7.11 shows one frameof the recorded movie, showingvalves, pistons, piston rods, pis-ton pins and piston rings. Ofspecial interest is the visualiza-tion of the oil cooling of the pis-ton bottoms. Since the pistonsare only connected to the enginebody via the piston rings withvery low heat dissipation, a con-tinuous oil jet is directed from be-low at the piston bottoms, low-ering the piston temperature bymore than 200 degrees C. In themovie, the oil jet of 1-2 mm di-ameter is clearly visible. Aroundthe upper turning point of the pis-ton, the dome-like spread of theoil at the underside of the pistoncan be observed.

Outlook

Further measurements will bedone in 2004, using a single-piston diesel engine to examinethe injection process. As soon asFRM-II is operational, measure-ments will commence in Garch-ing as well, trading beam inten-sity for better collimation. Fu-ture measurements will focus es-pecially on the oil cycle for thelubrication of the bearings of camshaft and crank shaft. To over-come the limitation in time res-olution given by the scintillator,new scintillation materials mustbe examined for feasibility forshort-time neutron radiography.The neutron quantum statisticsand the quantum noise of the im-age intensifier will be the finallimitation of achievable time res-olution, though it is by far notmet yet.

8 Events, People and Figures 91

8 Events, People and Figures

8.1 1st FRM-II Workshop on Neutron Scattering-Advanced Materials-21.7-24.7 Burg Rothenfels am Main

R. Georgii1, P. Böni2, A. Meyer 3, W. Petry 1, R. Röhlsberger3

1 ZWE FRM-II, TU München2 Physik-Department E21, TU München3 Physik-Department E13, TU München

Resume

On 21.7-24.7 the 1st FRM-IIWorkshop on Neutron Scatter-ing –Advanced Materials– tookplace at the Burg Rothenfels amMain. About 45 participantsmainly from the FRM-II instru-mentation group and the insti-tutes E13 and E21 of the PhysicsDepartment, but also guests fromas far as Australia (Prof. Stew-

ard Campbell) enjoyed the veryromantic scenery of the medi-vial castle above the picturesqueMain valley. This exceptionalsurrounding, together with thesuperb food and the beautifulBurgkeller of the castle led to ex-tensive scientific discussions be-sides the more then 30 talks givenduring the three days. A verybroad range of talks from variousfields as magnetism over viscous

melts to nuclear and fundamen-tal physics demonstrated the veryrich field for the use of the neu-trons as probes from large scalesto microscopic scales. Accord-ing the participants this work-shop helped to get an overviewof the science, which will be per-formed at the FRM II and shouldbe repeated in regular terms.

92 8 Events, People and Figures

8.2 Public Relations

V. Klamroth 1

1 Technische Universität München, Presse & Kommunikation, Bereich Garching

��

��

��

��

��

�

��

�

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��!��

"!#��

$��

The major event in 2003 wasthe visit of the Bavarian PrimeMinister Dr. Edmund Stoiber giv-ing the official starting signal forthe FRM-II in presence of theBavarian Minister of Science andthe Minister of EnvironmentalAffairs on 4th of June. About 40journalists, photographers, cam-eramen and other media represen-tatives came to Garching to coverthis event. Already the officialpermission of the Federal Min-istry of Environmental Affairs inthe middle of April had had alarge attraction for the media. Nu-merous requests from nationaland international journalists had

to be answered. This high mediainterest in the FRM-II last on forthe rest of the year. The mainquestions concerned the start ofoperation and the use of the neu-tron source. In addition therewere many requests for picturesand interviews with scientists.

The number of visitors inGarching was even higher than inthe year 2002. More than 4,500persons visited the FRM-II. Yetalone, 630 visitors came to theFRM-II at the open day in Octo-ber, a Saturday where all scien-tific institutes on the Campus inGarching were open to the pub-lic.

Staff of the FRM-II engagedwith dedication as guides all overthe year. This service to the pub-lic is even more creditable as thestrict access regulations to the re-actor and the long security proce-dures for visitors require a highlevel of patience.

Numerous politicians from dif-ferent parties including membersof the federal parliament weregathering information about theFRM-II and came to Garching.The large audience at the pub-lic lecture of Prof. Petry in theBavarian representation in Berlindemonstrated the high interest in


the new neutron source all overGermany.

The Bavarian Minister of Sci-ence showed the FRM-II to theMinister of Education of Luxem-bourg as an example of the suc-cessful Bavarian science politics.In a special issue of the engi-neer journal "Technik in Bayern"we presented the broad spectrumof the use of the new neutronsource.

As required by law, theleaflet "Rundherum sicher" ("Allaround safe") including arrange-ments in case of emergency wasdelivered to the inhabitants ofthe nearby villages and studentsand employees on the campusin August. The leaflet also de-scribes the safety of the reac-tor and shows by means of ap-propriate comparisons how lowthe radioactive release from thereactor is compared to naturalradioactivity. The German ver-sion of this leaflet can be orderedat the visitor’s office (Tel 289-12147) or directly downloadedfrom http://www.frm2.tum.de/publikationen. There isalso an English version available.

R u n d h e r u ms i c h e r !F o r s c h u n g s - N e u t ronenquelle FRM-II in Garching I n f o rmation für die Bevölkerung nach § 53 Strahlenschutzvero rd n u n g


8.3 Committees

Strategierat FRM-II

Chairman

Prof. Dr. Gernot HegerInstitut für KristallographieRWTH Aachen

Members

Ministerialdirigent Jürgen GroßkreutzBayrisches Staatsministerium für Wissenschaft,Forschung und Kunst

Dr. Rainer KöpkeBundesministerium für Bildung und Forschung

Prof. Dr. Georg BüldtInstitut für Biologische InformationsverarbeitungForschungszentrum Jülich

Prof. Dr. DoschMax-Planck-Institut für Metallforschung

Prof. Dr. Dieter RichterInstitut für FestkörperphysikForschungszentrum Jülich

Prof. Dr. Dirk SchwalmMax-Planck-Institut für Kernphysik, Heidelberg

Prof. Dr. Helmut SchwarzInstitute for ChemistryTechnische Universität Berlin

Prof. Dr. Dr. Michael WannenmacherRadiologische Klinik und PoliklinikAbteilung StrahlentherapieUniversität Heidelberg

Dr. Heinz VoggenreiterSenior ManagerCorporate Research and TechnologieEADS, Paris

Prof. Dr. Götz EckoldInstitute of Physical ChemistryUniversität Göttingen(Speaker Instrumentation advisory board)

Guests

Prof. Dr. Dr. Wolfgang A. HermannPräsident der Technischen Universität München

Dr.-Ing. Rainer KuchZentrale VerwaltungTechnische Universität München

Dr. Michael KlimkeZentrale VerwaltungTechnische Universität München

Prof. Dr. Winfried PetryZWE FRM-IITechnische Universität München

Prof. Dr. Klaus SchreckenbachZWE FRM-IITechnische Universität München

Guido EngelkeZWE FRM-IITechnische Universität München

Dr. Viola KlammrothZWE FRM-IITechnische Universität München


TUM advisory board

Chairman

Prof. Dr. Ewald WernerLehrstuhl für Werkstoffkunde und -mechanikTechnische Universität München

Members

Prof. Dr. Peter BöniPhysik Department E21Technische Universität München

Prof. Dr. Andreas TürlerInstitut für RadiochemieTechnische Universität München

Prof. Dr. Markus SchwaigerNuklearmedizinische Klinik und PoliklinikKlinikum rechts der IsarTechnische Universität München

Prof. Dr. Bernhard WolfHeinz Nixdorf-Lehrstuhl für medizinische Elek-tronikTechnische Universität München

Prof. Dr. Arne SkerraLehrstuhl für Biologische ChemieTechnische Universität München

Guests

Dr. Markus BusoldUniversity ManagementTechnische Universität München(until October 2003)

Dr. Michael KlimkeUniversity ManagementTechnische Universität München(since October 2003)


Prof. Winfried PetryZWE FRM-IITechnische Universität München

Prof. Klaus SchreckenbachZWE FRM-IITechnische Universität München


Instrumentation advisory board

Chairman

Prof. Dr. Götz EckoldInstitut for Physical ChemistryGeorg-August-Universität Göttingen

Members

Prof. Dr. Dieter HabsSektion PhysikLudwig-Maximilian-Universität München

Prof. Dr. Peter BöniPhysik Department E21Technische Universität München

Prof. Dr. Dirk DubbersPhysikalisches InstitutUniversität Heidelberg

Dr. Hans GrafAbteilung NEHahn-Meitner-Institut, Berlin

Prof. Dr. Andreas MagerlChair for Crystallography and Structural PhysicsFriedrich-Alexander-Universität Erlangen-Nürnberg

Prof.Dr. Gottfried MünzenbergGesellschaft für Schwerionenforschung mbH,Darmstadt

Prof. Dr. Wolfgang SchererLehrstuhl für Chemische PhysikUniversität Augsburg

Dr. habil. Dieter SchwahnInstitut für FestkörperforschungForschungszentrum Jülich

Prof. Dr. Markus SchwaigerLehrstuhl für NuklearmedizinTechnische Universität München

Prof. Dr. Andreas TürlerInstitut für RadiochemieTechnische Universität München

Dr. habil. Regine WillumeitGKSS Forschungszentrum, Geesthacht

Guests

Dr. Markus BusoldUniversity ManagementTechnische Universität München (until October2003)

Dr. Michael KlimkeZentrale VerwaltungTechnische Universität München (since Ocotber2003)

Dr. Klaus FeldmannBEO-PFRForschungszentrum Jülich

Ministerialdirigent Jürgen GroßkreutzBayrisches Staatsministerium für Wissenschaft,Forschung und Kunst

Prof. Dr. Gernot HegerInstitut für KristallographieRWTH Aachen

RR Felix KöhlBayerisches Staatsministerium für Wissenschaft,Forschung und Kunst


ORRin Birgit SchmidBayerisches Staatsministerium für Wissenschaft,Forschung und Kunst

Dr. Jürgen NeuhausZWE-FRM-IITechnische Universität München

Prof. Dr. Winfried PetryZWE-FRM-IITechnische Universität München


Prof. Dr. W. PressInstitut Laue LangevinGrenoble, France

Prof. Dr. Klaus SchreckbachZWE FRM-IITechnische Universität München

Prof. Dr. Tasso Springer

8.4 Staff

Board of Directors

Scientific directorProf. Dr. W. Petry

Technical directorProf. Dr. K. Schreckenbach

Administrative directorG. Engelke

Experiments

HeadProf. Dr. W. Petry

SecretariesE. DollW. Wittowetz

Figure 8.1:Experiments staff

CoordinationDr. J. Neuhaus

InstrumentsN. ArendH. Bamberger

M.BleuelJ. BrunnerK. BuchnerE. CalzadaD. Etzdorf


J. Franke (MPI-Stuttgart)Dr. R. GeorgiiDr. R. GillesF. GrünauerDr. E. GutsmiedlF. HibschDr. M. HofmannDr. M. Hölzel (TU Darmsatdt)Dr. K. Hradil (Univ. Göttingen)Dr. C. HugenschmidtDr. T. Keller (MPI-Stuttgart)Dr. V. KudryaschovB. Krimmer (TU-Darmstadt)D. LamagoDr. P. LinkR. LorenzT. MehaddeneDr. M. MevenM. MiseraM. MühlbauerA. MüllerDr. B. PedersenR. RepperJ. RingeDr. P. Rühm (MPI Stuttgart)Dr. B. SchillingerDr. M. Schlapp (TU Darmstadt)H. Schneider (Univ. Göttingen)

G. SeidlDr .T. UnruhH. WagensonnerDr. W. WaschkowskiDr. U. Wildgruber(MPI-Stuttgart)

Detectors and electronicDr. K. ZeitelhackS. EgerlandDr. A. KastenmüllerD. MaierM. Panradl

Sample environmentDr. J. PetersH. KolbA. PscheidtA. SchmidtS. SedlmairJ. Wetzlaff

Neutron opticsProf. Dr. G. BorchertC. BreunigH. HofmannE. KahleDr. S. MassalovitchC. Schanzer

A. Urban

Instrument control andsoftwareJ. KrügerJ. DettbarnS. GalinskiH. GildeS. PraßlS. Roth

Networking and computersJ. BeckmannJ. ErtlS. KrohnR. MüllerJ. MittermaierJ. PulzF. Zoebisch

Public relationsDr. W. WaschkowskiDr. V. Klamroth (ZV, TUM)K. SchaumlöffelJ. JeskeI. Maier

GuestsProf. S. Campbell (Univ. NSWAustralia)

Administration

HeadG. Engelke

SecretaryC. Zeller

MembersB. BendakB. GallenbergerI. HeinathH. NiedermeierR. Obermeier

Figure 8.2:Administration staff


Reactor operation

HeadProf. Dr. K. Schreckenbach

SecretariesK. LüttigM. NeubergerS. Rubsch

Visitor serviceJ. JeskeI. Maier

ManagementDr. H. Gerstenberg (irradiationand sources)Dr. J. Meier (reactor projects)Dr. C. Morkel (reactor operation)

Shift membersF. GründerA. BancsovA. BenkeM. DannerM. FlieherH. GroßL. Herdam

Figure 8.3:Reactor operation staff

K. HöglauerT. KalkG. KalteneggerU. KappenbergF. KewitzM. G. KrümpelmannJ. KundA. LochingerG. MauermannA. MeilingerM. MoserL. RottenkolberG. SchlittenbauerN. Wiegner

Technical servicesN. WaasmaierJ. AignerD. BahmetF. DollH. GampferW. GlashauserJ. GroßG. GuldF. Hofstetter

M. KleidorferW. KlugeH. KollmannsbergerJ.-L. KraußR. MaierK. OttoA. SchindlerR. Schlecht jun.J. SchreinerH. SedlmeierC. StroblG. WagnerA. WeberJ. WetzlC. Ziller

Electrics and electronicsR. SchätzleinG. AignerW. BuchnerÜ. SarikayaH. Schwaighofer


Irradiation and sourcesDr. H. GerstenbergJ.-M. FavoliDr. E. GutsmiedlDr. X. LiC. MüllerM. OberndorferD. PätheW. LangeG. LangenstückW. Waschkowski

Reactor projectsDr. J. Meier

DocumentationV. ZillJ. Jung

Instrument SafetyR. Lorenz

Technical designF.-L. TralmerJ. FinkH. FußstetterJ. JüttnerK. LichtensteinP. MrossM. Ullrich

WorkshopsC. HerzogU. StiegelA. BegicM. FußA. HuberA. ScharlR. Schlecht sen.

M. Tessaro

Radiation protectionDr. H. ZeisingS. DambeckW. DollrießDr. H. GerstenbergH. HottmannDr. J. MeierB. NeugebauerF. M. WagnerS. WolffH.-J. Werth

Chemical laboratoryDr. F. DienstbachT. AsamC. AuerR. Bertsch

Reactor physics

Prof. Dr. K. Böning Dr. A. Röhrmoser N. Wieschalla

Security department

L. StienenJ. Stephani


8.5 Publications

[1] Arend N., Gähler R., Keller T., Georgii R., Hils T., Böni P.Classical and quantum-mechanicalpicture of NRSE - Measuring the longitudinal Stern-Gerlach effect by means of TOF methods.Physics Letters A. Submitted.

[2] Asthalter T., Sergueev I., Franz H., Petry W., Messel K., R.Verbeni.Glass dynamics and scalingbehavior under pressure using quasielastic nuclear forward scattering. Hyperfine Interactions,C5(29-32).

[3] Attié D., Cordier B., Gros M., Laurent P., Schanne S., Tauzin G., von Ballmoos P., Bouchet L.,Jean P., Knödlseder J., Mandrou P., Paul P., Roques J.-P., Skinner G., Vedrenne G., Georgii R.,von Kienlin A., Lichti G., Schönfelder V., Strong A., Wunderer C., Shrader C., Sturner S.,Teegarden B., Weidenspointner G., Kiener J., Porquet M.-G., Tatischeff V., Crespin S., Joly S.,André Y., Sanchez F., Leleux P.INTEGRAL/SPI ground calibration. Astron. & Astrophys., 411,(2003), L71–L79.

[4] Bader A., Faschinger W., Hradil K., Kumpf C., Schallenberg T., Schumacher C., Geurts J.,Neder K., Molenkamp L. W.Two-step relaxation of ZnSe grown on (001) GaAs. J. Appl. Phys.In print.

[5] Bleuel M., Demmel F., Gähler R., Golub R., Habicht K., Keller T., Klimko S., Köper I.,Longeville S., Prokudaylo S. In: Mezei F., Pappas C., Gutberlet T. (Editors),Neutron-spin-echo-spectroscopy, Lecture Notes in Physics (Springer, 2003).

[6] Böning K., Waschkowski W., Schreckenbach K.Konzept und Sicherheit des FRM-II. Technik inBayern, 4, (2003), 12 – 13.

[7] Böning K., Neuhaus J.The Neutron Beams of TUM’s FRM-II. To appear in the Compendium onPurpose-Designed Research Reactor Features of the International Atomic Energy Agency (IAEA),Vienna, manuscript 9 pages.

[8] Brunner J., Frei G., Lehmann E., Vontobel P., Schillinger B.Steps towards dynamic neutronradiography of a combustion engine. PSI Scientific Report, III , (2003), 164.

[9] Cadogan J. M., Ryan D. H., Moze O., Suharyana, Hofmann M.Magnetic ordering in ErFe6Sn6. J.Phys.: Condens. Matter., 15, (2003), 1757.

[10] Chatake T., Ostermann A., Kurihara K., Parak F. G., Niimura N.Hydration in proteins observedby high-resolution neutron crystallography. Proteins, 50, (2003), 516–523.

[11] Ehrhardt H., Campbell S. J., Hofmann M.Magnetism of the nanostructured spinel zinc ferrite.Scripta Materialia, 2003, (2003), 1141.

[12] Engler N., Ostermann A., Niimura N., Parak F. G.Hydrogen atoms in proteins: Position anddynamics. Proc. Natl. Acad. Sci. USA, 100, (2003), 10243–10248.

[13] Engler N., Prusakov V., Ostermann A., Parak F.A water network within a protein: temperature-dependent water ligation in H64V-metmyoglobin and relaxation to deoxymyoglobin. Eur. Bio-phys. J., 31, (2003), 595–607.

[14] Faestermann T., Assmann W., Beck L., Bongers H., Carli W., Groß M., Großmann R., Habs D.,Hartung P., Heinz S., Jüttner P., Kester O., Krücken R., Maier H.-J., Maier-Komor P., Nebel F.,Ospald F., Pasini M., Rudolph K., Schumann M., Szerypo J., Thirolf P. G., Tralmer F., Zech E.The Munich Accelerator for Fission Fragments (MAFF). In: Proc. 6th Int. Conf. on RadioactiveNuclear Beams (RNB6)(Argonne, Illinois, 2003). To be published in Nucl. Phys. A.

[15] Frank A. I., Balashov S. N., Bondarenko I. V., Geltenbort P., Høghøj P., Masalovich S. V.,Nosov V. G. Phase modulation of a neutron wave and diffraction of ultracold neutrons on amoving grating. Physics Letters A, 311, (2003), 6–12.

[16] Frey F., Weidner E., Pedersen B.Tieftemperaturmessungen zur Strukturuntersuchung an dekago-nal en Quasikristallen. In preparation.


[17] Georgii R., Böni P., Pleshanov N.Polarised reflectometry with MIRA at the FRM-II. Physica B:Cond. Matt., 335, (2003), 250–254.

[18] Georgii R., Böni P., Pleshanov N.The VCN reflectometer for MIRA at the FRM-II. Langumir, 19,(2003), 7794–7795.

[19] Georgii R., Böni P., Lamago D., Stuber S., Grigoriev S., Malyeev S.Critical scattering of MnSi.Physica B. Accepted.

[20] Gilles R., Mukherji D., Genovese D. D., Strunz P., Barbier B., Kockelmann W., Rösler J., Fuess H.Misfit Investigations of Nickel-base Superalloys. Materials Science Forum, 426-432, (2003),821.

[21] Habicht K., Keller T., Golub R.The resolution function in neutron spin echo spectroscopy withthree axis spectrometers. J. Appl. Cryst., 36, (2003), 1307.

[22] Habicht K., Golub R., Keimer B., Mezei F., Keller T.Phonon lifetime measurements with neutronresonance spin echo at BENSC - anharmonic lattice dynamics. In: Proceedings SCNS(ILL,2003).

[23] Habicht K., Golub R., Keller T.Resolution issues for neutron spin echo spectroscopy on tripleaxis spectrometers. In: Proceedings SCNS(ILL, 2003).

[24] Habicht K., Golub R., Gähler R., Keller T. In: Mezei F., Pappas C., Gutberlet T. (Editors),Neutron-spin-echo-spectroscopy, Lecture Notes in Physics (Springer, 2003).

[25] Habs D., Groß M., Assmann W., Ames F., Bongers H., Emhofer S., Heinz S., Henry S., Kester O.,Neumayr J., Ospald F., Reiter P., Sieber T., Szerypo J., Thirolf P., Varentsov V., Wilfart A.,Faestermann T., Krücken R., Maier-Komor P.The Munich Accelerator for Fission FragmentsMAFF. Nucl. Instr. and Meth. B, 204.

[26] Herrmann R. A., Pedersen B., Scherer W., Shorokhov D.Hyperconjugative Delocalization inSilatranes. Submitted to Chem. Comm.

[27] Hoelzel M., Danilkin S., Ehrenberg H., Toebbens D. M., Udovic T. J., und H. Wipf H. F.Effects ofhigh-pressure hydrogen charging on the structure of austenitic stainless steels. Mat. Sci. & Eng.A. Submitted.

[28] Hoelzel M., Gilles R., Schlapp M., Boysen H., Fuess H.Monte Carlo simulations of variousinstrument configurations of the new Structure Powder Diffractometer (SPODI). Physica B. Inprint.

[29] Hofmann M.Zerstörungsfreie Materialprüfung. Technik in Bayern, 7. Jhg(4), (2003), 18.[30] Hofmann M., Campbell S. J., Kaczmarek W. A., Welzel S.Mechanochemical Transformation of

α-Fe2O3 to Fe3−xO4 - Microstructural Investigation. J. Alloys Comp., 348, (2003), 278.[31] Hofmann M., Campbell S. J., Link P., Goncharenko I., Knorr K.Pressure Induced Magnetic and

Valence Transition in YbMn2Ge2. Physica B. In press.[32] Hofmann M., Campbell S. J., Edge A. V. J.Valence and magnetic transitions in YbMn2Si2−xGex.

J. Mag. Mag. Mater.In press.[33] Hugenschmidt C., Straßer B., Schreckenbach K.Investigation of the Annealed Copper Surface

by Positron Annihilation Induced Auger Electron Spectroscopy. Radiat. Phys. Chem., 68(3-4),(2003), 627–629.

[34] Hugenschmidt C., Liszkay L., Egger W.Investigation of Laser Treated AlN by Positron Life-time Measurements. In: Proceedings of the 3rd International Workshop on Positron Studies ofSemiconductor Defects PSSD-2002(Sendai, Japan, 2003).

[35] Hugenschmidt C., Kögel G., Repper R., Schreckenbach K., Sperr P., Triftshäuser W.PositronSource Based on Neutron Capture. Radiat. Phys. Chem., 68(3-4), (2003), 669–671.

[36] Hugenschmidt C., Kögel G., Repper R., Schreckenbach K., Sperr P., Straßer B., Triftshäuser W.NEPOMUC - The New Positron Beam Facility at FRM II. Mat. Sci. Forum. In press.

[37] Kardjilov N., S.Baechler, Basturk M., Schillinger B., et al.New features in cold neutron radiog-


raphy and tomography - Part II: applied energy-selective neutron radiography and tomography.Nucl. Instrum. Meth. A, 501, (2003), 536–546.

[38] Keller T., Keimer B., Habicht K., Golub R., Mezei F. In: Mezei F., Pappas C., Gutberlet T.(Editors),Neutron-spin-echo-spectroscopy, Lecture Notes in Physics (Springer, 2003).

[39] Korsounski V., Neder R., Hradil K., Barglik-Chory C., Müller G., Neuefeind J.Investigationof nanocrystalline CdS-Glutathione particles by radial distribution function. J. Appl. Cryst., 36,(2003), 1389–1396.

[40] Lamago D., Dameris M., Schnadt C., Eyring V., Brühl C.Impact of large solar zenith angles onlower stratospheric dynamical and chemical processes in a coupled chemistry-climate model.Atmos. Chem. Phys., 3, (2003), 1981–1990.

[41] Lehmann E., Frei G., Hartmann S., Vontobel P., Brunner J., Schillinger B., Abele H., Gildemeis-ter A., Gähler R. Dynamic imaging with a triggered and intensified system at a high intenseneutron beam. PSI annual report 2003. In print.

[42] Liszkay L., Kögel G., Sperr P., Egger W., Hugenschmidt C., Triftshäuser W.Positron BeamSplitter at the High Intensity Positron Beam in Munich. Mat. Sci. Forum. In press.

[43] Longeville S., Doster W., Diehl M., Gähler R., Petry W.Neutron Resonance Spin-Echo: Oxygentransport in crowded protein solutions. Lecture Notes in Physics, 601, (2003), 325–335.

[44] Major J., Dosch H., Felcher G. P., Habicht K., Keller T., te-Velthuis S. G. E., Vorobiev A.,Wahl M. Combining of neutron spin echo and reflectivity: a new technique for probing surfaceand interface order. Physica B, 336, (2003), 8.

[45] Molls M., Kneschaurek P., Waschkowski W.Tumortherapie mit schnellen Neutronen. Technik inBayern, (4).

[46] Moze O., Hofmann M., Cadogan J. M., Buschow K. H. J., Ryan D. H.Magnetic order in RCr2Si2intermetallics. European Physical Journal B. In press.

[47] Mukherji D., Strunz P., Genovese D. D., Gilles R., Rösler J.Investigation of microstructuralchanges in Inconel 706 at high temperatures by in situ SANS. Metallurgical Transaction, 34,(2003), 2781.

[48] Mukherji D., Gilles R., Barbier B., Genovese D. D., Hasse B., Strunz P., Wroblewski T.,Fuess H., Rösler J.Lattice misfit measurements in INCONEL 706 containing coherentγ ′ and γ ′′

precipitates. Scripta Materialia, 48, (2003), 333.[49] Neder R. B., Korsounski V., Hradil K., Barglik-Chory C., Müller G.Defect structure of glutathione

stabilized CdS nanoparticles. J. Appl. Cryst.Accepted for publication.[50] Niimura N., Chatake T., Ostermann A., Kurihara K., Tanaka T.High resolution neutron protein

crystallography. Hydrogen and hydration in proteins. Z. Kristallogr., 218, (2003), 96–107.[51] Pasini M., Kester O., Habs D., Groß M., Sieber T., Maier H.-J., Assmann W., Krücken R.,

Faestermann T., Schempp A., Ratzinger U., Safvan C. P.Radioactive Ion Beam Acceleration atMAFF. In: Proc. 6th Int. Conf. on Radioactive Nuclear Beams (RNB6)(Argonne, Illinois, 2003).To be published in Nucl. Phys. A.

[52] Petry W.Advanced Neutron Instrumentation at FRM-II. atw 48, 5, (2003), 315–318.[53] Petry W.Neutronen für Lebensqualität. Technik in Bayern, 4, (2003), 8–10.[54] Rajevac V., Hoelzel M., Danilkin S. A., und H. Fuess A. H.Phonon dispersion in austenitic stain-

less steels Fe-18Cr-12Ni-2Mo and Fe-18Cr-16Ni-10Mn. J. Phys. C: Cond. Matter.Submitted.[55] Rekveldt M. T., Bouwman W. G., Kraan W. H., Uca O., Grigoriev S. V., Habicht K., Keller T.

In: Mezei F., Pappas C., Gutberlet T. (Editors),Neutron-spin-echo-spectroscopy, Lecture Notes inPhysics (Springer, 2003).

[56] Roth S., Burghammer M., Gilles R., Mukherji D., Rösler J., Strunz P.Precipitate scanning inNi-baseγ /γ ′ superalloys. NIM B, 200, (2003), 255.

[57] Schillinger B. Neutronen sehen, was Röntgenstrahlen verborgen bleibt. Technik in Bayern, 4,


(2003), 14–15.[58] Schlapp M., Hoelzel M., Gilles R., Ioffe A., Brueckel T., Fueß H., von Seggern H.Neutron image

plates with lowγ-sensitivity. In: Proceedings of ECNS(2003). Submitted to Physica B.[59] Schlapp M., Bulur E., von Seggern H.Photo-stimulated luminescence of calcium co-doped

BaFBr:Eu2+ x-ray storage phosphors. J. Phys. D: Appl. Phys., 36, (2003), 103–108.[60] Schmidt M., Hofmann M., Campbell S. J.Magnetic Structure of Strontium Ferrite Sr4Fe4O11. J.

Phys.: Condens. Matter., 15, (2003), 8691.[61] Stampanoni M., Borchert G., Abela R., Rüegsegger P.A Bragg Magnifier With Sub-µm Resolution

Using High Energy Synchrotron Light. In: AIP Conf. Proc., volume 652, 112 (2003).[62] Stampanoni M., Borchert G., Abela R., Rüegsegger P.Nanotomography based on double

asymmetrical Bragg diffraction. Applied Physics Letters, 82, (2003), 2922–2924.[63] Stampanoni M., Borchert G., Abela R., Rüegsegger P.Nanotomography based on double

asymmetrical Bragg diffraction. Virtual Journal of Nanoscale, Science & Technology, 7(18).[64] Stampanoni M., Borchert G., Rüegsegger P., Abela R.Submicrometer X-ray tomography with

double asymmetrical Bragg diffraction. PSI Scientific Report Annex VII, 52–54.[65] Strunz P., Gilles R., Mukherji D., Wiedenmann A.Anisotropic SANS data evaluation: a faster

approach. J. Appl. Cryst., 36(854).[66] Strunz P., Mukherji D., Gilles R., Rösler J., Wiedenmann A.Small-Angle Neutron Scattering: a

Tool for Microstructural Investigation of High-Temperature Materials. Materials Science Forum,426 - 432, (2003), 755.

[67] Szytula A., Penc B., Jezierski A., Hofmann M., Campbell S. J.Electronic Structure of YbMn2X2

(X = Si,Ge) Compounds. Acta Phys. Pol., 34, (2003), 1561.[68] Waschkowski W., Lange W., Böning K.Fast Neutrons for Tumor Treatments and Technical

Applications at the FRM-II. To appear in the Compendium on Purpose-Designed ResearchReactor Features of the International Atomic Energy Agency (IAEA), Vienna, manuscript 9 pages.

[69] Weidner E., Pedersen B., Frey F., C. P.High-resolutuion x-ray studies on martian pyroxenes. Inpreparation.

[70] Wronkowska A., Wronkowski A., Bukalek A., Stefanski M., Arwin H., Firszt F., Legowski S.,Meczynska H., Hradil K.Investigation of Cd1− xMnxTe crystals by means of ellipsometry andauger electron spectroscopy. Applied Surface Science, 212, (2003), 110–115.

[71] Wu E., Campbell S. J., Kaczmarek W. A., Hofmann M., Kennedy S.Nanostructured(CoxFe1−x)3−yO4 Spinel - Mechanochemical Synthesis. Z. Metallkunde. In press.

[72] Wurst K., Schweda E., Bevan D. J. M., Mohyla J., Wallwork K. S., Hofmann M.Single-crystalstructure determination of Zr50Sc12O118. Solid State Science, 15, (2003), 1491.

[73] Zeising H.FRM-II: Uses of the neutron source for research, medicine and industry. Conferencetranscript of the 5th international symposium on radiation protection of the TÜV academy (2003).

[74] Zeising H.-W. Internal radiation protection at the "neutron research source FRM-II". Confer-ence transcript of the 12th expert talk "Monitoring the environmental radiation".


Imprint

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Publisher:Technische Universität MünchenNeue ForschungsneutronenquelleZWE-FRM-IILichtenbergstr. 185747 GarchingGermanyPhone: +49 89-289-14965Fax: +49 89-289-14995Internet: http://www.frm2.tum.deemail: [email protected]

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Editors:J. NeuhausB. Pedersen

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Photographic credits:All images: TUM

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Design:B. Pedersen, TUMJ. Neuhaus, TUMC. Simon

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Typesetting(LATEX 2ε ):B. Pedersen, TUM

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

annual report 2003 zwe frm-ii · 2015-03-04 · frm-ii takes part and will welcome its ﬁrst eu...

Documents