ASAP 2002
Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources Editors
Paul E. Koehler
Christopher R. Gould
Robert C. Haight
Timothy E. Valentine
Symmetry Experiments
Proposed ASAP Beam Line at SNS
Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources
This page is intentionally left blank
ASAP 2002 Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources Editors
Paul E. Koehler Oak Ridge National Laboratory, USA
Christopher R. Gould North Carolina State University, USA
Robert Haight Los Alamos National Laboratory, USA
Timothy E. Valentine Oak Ridge National Laboratory, USA
V f e World Scientific wll
New Jersey • London Singapore • Si • Hong Kong
Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: Suite 202, 1060 Main Street, River Edge, NJ 07661 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE
British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.
ASTROPHYSICS, SYMMETRIES, AND APPLIED PHYSICS AT SPALLATION NEUTRON SOURCES Copyright © 2002 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.
For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.
ISBN 981-238-249-6
Printed by Fulsland Offset Printing (S) Pte Ltd, Singapore
FOREWORD AND ACKNOWLEDGMENTS Intense beams of epithermal neutrons from spallation sources are enabling the exploration of new vistas of research in nuclear astrophysics, the study of violation of fundamental symmetries, and applied nuclear physics. There has been, and continues to be an active and productive world-wide community engaged in these areas of research at existing facilities such as electron linacs (e.g. at the Oak Ridge Electron Linear Accelerator, ORELA) and van de Graaff laboratories (e.g. in Karlsruhe, Germany and at the Tokyo Institute of Technology in Japan) as well as new efforts at spallation sources (e.g. at the Los Alamos Neutron Science Center, LANSCE and the CERN n_TOF facility). Building on, the long and distinguished history of research at these facilities, the much higher flux available at spallation sources makes it possible to work with much smaller samples as well as to make measurements with much higher precision. These new capabilities of spallation sources are enabling the exploration of new, exciting areas of research such as: • The nucleosynthesis of the elements in dynamic stellar environments such as pulsing red giant stars and supernovae. • The study of fundamental symmetries such as the nature of parity violation in complex nuclei and the search for violations of time-reversal invariance. • The feasibility and efficiency of using accelerator-driven sub-critical assemblies to transmute dangerous, long-lived radioactive waste to more benign materials as well as several other topics in applied nuclear physics such as criticality safety and nuclear medicine. A unifying feature of all of these fields is the need for the highest intensity source of pulsed epithermal neutrons. The new Spallation Neutron Source (SNS) being built at Oak Ridge National Laboratory (ORNL) will be by far the highest flux pulsed source of epithermal neutrons in the world when it comes on line in 2006. Although the main thrust of the science program at the SNS will be materials science, the facility could provide outstanding opportunities for research in nuclear astrophysics, fundamental symmetries, and applied nuclear physics. To review the current status of these fields and to begin to assemble the scientific case and the community of researchers for future experiments at the SNS, a workshop on "Astrophysics, Symmetries, and Applied Physics" was held during March 11-13 at ORNL. Over 60 scientists, representing 11 different US and 4 different foreign universities as well as many national laboratories around the world participated in the workshop. We thank the speakers for their excellent presentations and everyone for participating in the discussions and making the workshop a success. The scientific organizing committee would like to thank Fred E. Bertrand, Jr. (ORNL Physics Division), John C. Nemeth (Oak Ridge Associated Universities, ORAU), and R. Gil Gilliland (ORNL) for financial support to make the workshop
v
vi possible. We would also like to thank Doro Wiarda (ORNL, Physics Division) for setting up the web site for the workshop. Finally, we would like to thank Ken Carter and staff of ORAU for providing technical and administrative support to the collaboration and to Carlene Stewart (ORAU) for her excellent help in organizing and running the workshop. Paul Koehler Chris Gould Bob Haight Tim Valentine
Schedule for the Workshop on Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources
Sunday March 10
19:00-21:00
Reception at Pollard Auditorium in Oak Ridge
Monday March 11
Welcome (Chair: G. Young) 9:00-9:10 9:10-9:15 9:15-9:30
Welcome/G. Young/ORNL Welcome/R. Townsend/ORAU Welcome/J. Roberto/ORNL
Nuclear Astrophysics (Chair: G. Young) 9:30-10:15 10:15-10:45
Laboratory Experiments for Neutron Capture Nucleosynthesis! F. Kappeler, FZK Karlsruhe, Germany DANCE at LANSCE J. Ullmann, LANL
10:45-11:00
Break
11:00-11:30 11:30-12:00
The Astrophysics Program at the CERN n_TOF Facility A. Mengoni, ENEA Bologna, Italy Recent Astrophysics Results from ORELA and Possible Future Experiments at ORELA and SNS P. Koehler, ORNL
12:00-13:30
Lunch
Nuclear Astrophysics (continued) (Chair: Art Champagne) 13:30-13:50
13:50-14:10
Sensitivity of Isotope Yields to Reaction Rates in the Alpha-Rich Freezeout C. Jordan, Clemson University Neutron Reactions of Light Nuclei from Astrophysics and Nuclear Physics Interest
VIII
14:10-14:30 14:30-14:50
14:50-15:10
15:10-15:30
Y. Nagai, Osaka University, Japan "Offline" Radioactive Targets R. Rundberg, LANL Measurement ofthep(n, y)d Cross Section for Big Bang Nucleosynthesis at the Spallation Source LANSCE E. Esch, LANL Phonon Properties of Materials from Resonance Doppler Broadening J. E. Lynn, LANL Break
Applied Nuclear Physics (Chair: R. Haight) 15:30-16:15 16:15-16:45
Applied Physics at Spallation Neutron Sources P. Oblozinsky, BNL Applied Physics Measurements at the CERN n_TOF Facility E. Gonzalez-Romero, CIEMAT Madrid, Spain
Tuesday March 12
Applied Nuclear Physics (continued) (Chair: P. Oblozinsky) 9:00-9:30
Activities of the DOE Nuclear Criticality Safety Program (NCSP) at the Oak Ridge Electron Linear Accelerator (ORELA) T. Valentine, ORNL
9:30-9:50
Parameters for Nuclear Reaction Calculations -Needs for Improvement M. Herman, IAEA Vienna, Austria New Al(n, tf Measurements and Criticality Safety L. Leal, ORNL
9:50-10:10
10:10-10:30
Break
10:30-10:50
Time-of Flight Spectrometer GNEIS with Spallation Neutron Source A. Laptev, PNPI, Gatchina, Russia Measurements of Neutron Capture Cross Sections of Long-Lived Fission Products H. Harada, JNC, Japan
10:50-11:10
IX
11:10-11:30
11:30-11:50
11:50-13:30
Temperature Measurements in Dynamically-Loaded Systems Using Neutron Resonance Spectroscopy at LANSCE V. Yuan, LANL Radioactive Targets from RIA J. Blackmon, ORNL Lunch
Symmetries (Chair: C. Gould) 13:30-14:00 14:00 -14:30
14:30-14:50 14:50-15:10
15:10-15:30
Parity Violation in Neutron Resonances G. Mitchell, North Carolina State University New Experimental Capabilities for Parity Non-Conservation and Time Reversal InvarianceViolation S. Penttila, LANL Symmetry Tests at the Japanese Spallation Neutron Source Y. Masuda, KEK, Japan Violation of Fundamental Symmetries in Resonance Neutron Induced Fission W. Furman, JINR Dubna, Russia Physics of the Fission Process and Parity Violation in Neutron Induced Reactions V. Gudkov, University of South Carolina
15:30-15:50
Break
15:50-16:20
New Possibilities for Parity Violation Studies A. Hayes, LANL Time Reversal Tests and Sign Correlations in Heavy Nuclei C. Gould, North Carolina State University A Five-Fold Correlation Experiment to Measure Time Reversal Invariance Violation Using Neutron Resonances in Holmium P. Huffman, NIST
16:20-16:40 16:40-17:00
X
Wednesday March 13
SNS and Discussions (Chair: Paul Koehler) 9:00-9:30 9:30-10:15
SNS Letter of Intent and Instrument Development Team Processes T. Mason, ORNL SNS Technical Overview P. Ferguson, ORNL
10:15-10:30
Break
10:30-12:00
Discussion of SNS Beam Line Requirements, Possible Experimental Program, and LoI/IDT Issues Discussion leader: P. Koehler, ORNL
ASAP 2002 PARTICIPANT LIST Name
Affiliation
Email
Chuck Alexander
Oak Ridge National Laboratory
[email protected]
Dan Bardayan
Oak Ridge National Laboratory
[email protected]
Jon Batchelder
Oak Ridge Associated Universities
[email protected]
Jeff Blackmon
Oak Ridge National Laboratory
[email protected]
Carl Brune
Ohio University
[email protected]
Ken Carter
Oak Ridge Associated Universities
[email protected]
Art Champagne
University of North Carolina
[email protected]
Walter Furman
JINR, Dubna, Russia
[email protected] inr.ru
Kazuyoshi Furutaka
Japan Nuclear Cycle Development Inst.
[email protected]
T. Vincent Cianciolo
Oak Ridge National Laboratory
[email protected]
Aaron Couture
University of Notre Dame
[email protected]
Yaron Danon
Rensselaer Polytechnic Institute
[email protected]
Felix Difilippo
Oak Ridge National Laboratory
[email protected]
Ernst Esch
Los Alamos National Laboratory
[email protected]
Phillip D. Ferguson
Oak Ridge National Laboratory
[email protected]
Enrique Gonzalez
CIEMAT
[email protected]
Christopher Gould
North Carolina State University
[email protected]
Geoffrey Greene
Los Alamos National Laboratory
[email protected]
Colorado School of Mines
[email protected]
Uwe Greife Vladimir Gudkov Gueorgui Gueorguiev Robert Haight Hideo Harada Jack Harvey Anna Hayes Michal Herman Paul R. Huffman Masayuki Igashira Cal Jordan Franz Kaeppeler Guinyun Kim Paul Koehler
University of South Carolina
[email protected]
University of Florida
george@serverl .nuceng.ufi.edu
Los Alamos National Laboratory
[email protected]
Japan Nuclear Cycle Development Inst.
[email protected]
Oak Ridge National Laboratory
[email protected]
Los Alamos National Laboratory
[email protected]
International Atomic Energy Agency
herman@ndsalpha. iaea.org
Natl. Institute of Standards and Tech.
[email protected]
Tokyo Institute of Technology
[email protected]
Clemson University
[email protected]
FZK, Karlsruhe
[email protected]
Kyungpook National University
[email protected]
Oak Ridge National Laboratory
[email protected]
XII
Raymond Kozub
Tennessee Technological University
[email protected]
Alexandre Laptev
Petersburg Nuclear Physics Institute
[email protected]
Luiz Leal
Oak Ridge National Laboratory
[email protected]
Eric Lynn
Los Alamos National Laboratory
eric,
[email protected]
Thorn Mason
Oak Ridge National Laboratory
[email protected]
Yasuhiro Masuda
High Energy Accelerator Research Org.
[email protected]
Alberto Mengoni
ENEA-Applied Physics Division
[email protected]
Gary Mitchell
North Carolina State University
[email protected]
Paul Mueller
Oak Ridge National Laboratory
[email protected]
Yasuki Nagai
Osaka University
[email protected]
John Neal
Oak Ridge National Laboratory
[email protected]
Pavel Oblozinsky
Brookhaven National Laboratory
[email protected]
Toshiro Osaki
Tokyo Institute of Technology
[email protected]
Peter Parker
Yale University
[email protected]
Seppo Penttila
Los Alamos National Laboratory
[email protected]
S. Raman
Oak Ridge National Laboratory
[email protected]
Wolfgang Rapp
University of Karlsruhe
[email protected]
John-Paul Renier
Oak Ridge National Laboratory
[email protected]
James B. Roberto
Oak Ridge National Laboratory
[email protected]
Bob Rundberg
Los Alamos National Laboratory
[email protected]
Royce Sayer
Oak Ridge National Laboratory
[email protected]
Kenneth Toth
Oak Ridge National Laboratory
[email protected]\
Ronald Townsend
Oak Ridge Associated Universities
to wnsenr@orau. gov
John Ullmann
Los Alamos National Laboratory
[email protected]
Timothy Valentine
Oak Ridge National Laboratory
[email protected]
Steve Wender
Los Alamos National Laboratory
[email protected]
Jerry Wilhelmy
Los Alamos National Laboratory
[email protected]
Glenn Young
Oak Ridge National Laboratory
[email protected]
Vincent Yuan
Los Alamos National Laboratory
[email protected]
xiii
CONTENTS Preface Schedule for the Workshop on ASAP Participant List Neutron Capture Nucleosynthesis: Astrophysical Processes and Laboratory Approaches F. Kappeler The Detector for Advanced Neutron Capture Experiments at LANSCE J.L. Ullmann, R.C. Haight, L. Hunt, E. Seabury, R.S. Rundberg, J.B. Wilhelmy, MM. Fowler, D.D. Strottman, F. Kaeppeler, R. Reifarth, M. Heil and E.P. Chanberlin Astrophysics Program at the CERN n_TOF Facility A. Mengoni
v vii xi
1
16
25
Recent Astrophysics Results from ORELA and Possible Future Experiments at ORELA and SNS P.E. Koehler
32
Sensitivity of Isotope Yields to Reaction Rates in the Alpha Rich Freezeout G.C. Jordan IV and B.S. Meyer
42
Neutron Reactions of Light Nuclei from Astrophysics & Nuclear Physics Interest Y. Nagai, T. Shima, A. Tomyo, H. Makii, K. Mishima, M. Segawa, M. Igashira and T. Ohsaki
52
XIV
Measurement of the n+p—>d+y Cross Section for Big Bang Nucleosynthesis with the Spallation Neutron Source at the Los Alamos Neutron Science Center E.-I. Eschm, J.M. O'Donnell, S.A. Wender, D. Bowman, G. Morgan and J. Matthews
58
Phonon Properties of Materials from Neutron Resonance Doppler Broadening Measurements J. Eric Lynn
65
Applied Nuclear Physics at Spallation Neutron Sources Pavel Oblozinsky
73
Applied Physics Measurements at the CERN n_TOF(1) Facility E. Gonzalez
83
Activities of the DOE Nuclear Criticality Safety Program (NCSP) at the Oak Ridge Electron Linear Accelerator (ORELA) Timothy E. Valentine, Luiz C. Leal and Klaus H. Guber Parameters for Nuclear Reaction Calculations - Needs for Improvements M. Herman Aluminum Data Measurements and Evaluation for Criticality Safety Applications L.C. Leal, K.H. Guber, R.R. Spencer, H. Derrien and R.Q. Wright Nuclear Physics Investigations at the Time-of-Flight Spectrometer GNEIS with Spallation Neutron Source O.A. Shcherbakov, A.B. Laptev andA.S. Vorobyev
97
107
115
123
XV
Measurement of Neutron Capture Cross Sections of Long-lived Fission Products H. Harada, S. Nakamura, K. Furutaka, T. Katoh, M.M.H. Miah, O. Shcherbakov, H. Yamana, T. Fujii and K. Kobayashi Temperature Measurements in Dynamically-loaded Systems Using Neutron Resonance Spectroscopy (NRS) atLANSCE V.W. Yuan
131
138
Radioactive Target Production at RIA J.C. Blackmon
146
Parity Violation inEpithermal Neutron Resonances G.E. Mitchell, J.D. Bowman, S.I. Penttila and E.I. Sharapov
155
New Experimental Capabilities for Parity Non-conservation and the Time Reversal Invariance Violation in Neutron Transmission S.I. Penttila T-Violating Three-fold Correlation in Neutron Transmission Y. Masuda
164
175
Violation of Fundamental Symmetries in Resonance Neutron Induced Fission A. Barabanov, W. Furman and A. Popov
184
Physics of the Fission Process and Parity Violation in Neutron Induced Reactions Vladimir Gudkov
194
XVI
Possibilities for Studies of Parity Violation at the SNS Using the Capture Gamma Reaction A. C. Hayes and Luca Zanini Time Reversal Tests with Epithermal Neutrons C.R. Gould An Experiment to Search for Parity-conserving Time Reversal Invariance Using Epithermal Neutrons from the Spallation Neutron Source P.R. Huffman Neutronic Characteristics of the Spallation Neutron Source P.D. Ferguson, E.B. lverson and F.X. Gallmeier Workshop Summary: Opportunities in Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources Paul Koehler, Christoper Gould, Robert Haight and Timothy Valentine
202
209
217
225
233
Letter of Intent to the Spallation Neutron Source
234
Author Index
245
N E U T R O N C A P T U R E NUCLEOSYNTHESIS: ASTROPHYSICAL PROCESSES A N D LABORATORY APPROACHES F. K A P P E L E R Forschungszentrum
Karlsruhe,
Institut fur Kernphysik, Postfach Karlsruhe, Germany E-mail:
[email protected]
3640,
D-76021
Neutron reactions are responsible for the formation of the elements heavier than iron. The corresponding scenarios relate to helium burning in Red Giant stars (s process) and to supernova explosions (r and p process). The status of the relevant neutron data for the various scenarios are briefly summarized, followed by an outline of the essential experimental techniques. The direct impact of laboratory results on the interpretation of the observed abundance patterns and their role as crucial tests for astrophysical models are illustrated by representative examples. The very high flux at spallation neutron sources provide a unique possibility for investigating numerous difficult and hitherto inaccessible cases, in particular cross sections of the important radioactive nuclei. In combination with advanced detector concepts these facilities provide a promising step towards a quantitative picture of galactic chemical evolution.
1
Introduction
A first clue for the origin of the chemical elements was obtained in the 1930ies by the analysis of carbonaceous chondrites, a class of primitive meteorites, which preserved the original composition of the protosolar nebula *. At about the same time, nuclear burning was identified as the stellar energy source 2 3 4 ' ' . However, it was not before 1952, when Merrill 5 discovered Tc lines in the spectra of red giant stars - an unstable element with isotopic half-lives much shorter than the stellar evolution time - that stellar nucleosynthesis was accepted as the origin of the chemical elements. The various aspects of this new field of Nuclear Astrophysics, i.e. the elemental composition of astronomical objects, the standard abundance distribution, the nucleosynthesis mechanisms, and the related nuclear physics, were eventually combined in the fundamental and seminal paper by Burbidge, Burbidge, Fowler, and Hoyle 6 . A comprehensive summary of the 40 years of progress in nucleosynthesis since B 2 FH was published recently by Wallerstein et al. 7 . Reviews on more specific topics can be found elsewhere 8>9,10,11,12,13,14,15
1
2
100 MASS NUMBER
150
Figure 1. The isotopic abundance distribution in the solar system (from Ref. 1 7 ).
2 2.1
The Observed Abundances The Solar System
Any nucleosynthesis model must be checked against observations. Originally, the composition of the solar system was considered a standard which can be reliably derived by spectroscopy of the photosphere and by meteorite analyses 16,17 p r o m this distribution (Fig. 1) the signatures of the dominant scenarios can be inferred, starting with the very large primordial H and He abundances from the Big Bang. The abundances of the rare elements Li, Be, and B, which are difficult to produce because of the stability gaps at A=5 and 8, but are easily burnt in stars, were mostly formed by spallation reactions induced by galactic cosmic rays. Stellar nucleosynthesis starts with the ashes of He burning, 12 C and 1 6 0 , which are partly converted to 14 N by the CNO cycle in later stellar generations. In subsequent stages of stellar evolution, the light elements up to the mass 40 to 50 region are produced by charged particle reactions during C, Ne, and O burning 7 . The corresponding yields show a strong preference for the most stable nuclei built from a-particles. This part of the distribution is strongly influenced by the Coulomb barrier, resulting in an exponential decrease with increasing atomic number Z. Ultimately, Si burning leads to such high temperatures and densities that nuclear statistical equilibrium is reached. Under these conditions matter is transformed into the most stable
3 nuclei around Fe, giving rise to the dominant maximum at A = 5 6 . Due to the increasing Coulomb barriers the abundances of all heavier nuclei up to the actinides are essentially shaped by neutron capture nucleosynthesis, leading to a fairly flat distribution characterized by the pronounced r and s maxima. These twin peaks are the signatures of the slow (s) and rapid (r) neutron capture processes discussed below. 2.2
Galactic
Evolution
While the solar abundance distribution is characteristic for most stars it represents just the average enrichment of the Galaxy 4.55 Gyr ago. The chemical evolution prior to this point has become an intense field of investigation. Spectral analysis of stellar atmospheres has become an ever refined source of information. W i t h the astonshing sensitivity of ground and satellite based telescopes extremely faint a n d / o r metal poor stars can be observed from the UV to the far IR, providing an almost complete element p a t t e r n of these objects 1 8 ' 1 9 . Likewise, chemically peculiar stars, which witness ongoing sprocess nucleosynthesis in their deep interiors, or the expanding supernova ejecta can be accessed in great detail as well. For more t h a n three decades, direct spectroscopic observations have been complemented by analyses of circumstellar dust grains from AGB stars or supernovae, which survived the homogenization in the protosolar cloud and are preserved as minute inclusions in meteorites n > 1 4 . T h e isotopic composition of these presolar grains clearly exhibit enrichments, which can be a t t r i b u t e d t o particular nucleosynthetic scenarios such as the s or r process. T h e wealth of new and exciting information on the chemical evolution of the Galaxy calls for an expanded and improved nuclear physics d a t a base, which is indispensable for the quantitative interpretation of these observations, and hence for understanding the history of the universe. 3
N e u t r o n Capture Scenarios
W h e n the concept of neutron capture nucleosynthesis was first formulated 6 the s and r processes were already identified as the mechanisms responsible for the sharp maxima in the abundance distribution. These mechanisms are illustrated in Fig. 2, which shows the respective reaction p a t h s in t h e chart of nuclides. T h e s process being characterized by relatively low neutron densities implies neutron capture times much longer t h a n typical /?-decay half-lives. Therefore, the s-process reaction p a t h follows t h e stability valley as indicated
Seed for s-Process
s-Process Reaction Path
s-Branchings ( M N i 79Se 8 5 Kr,...)
Figure 2. An illustration of the neutron capture processes responsible for the formation of the nuclei between iron and the actinides. The observed abundance distribution in the inset shows characteristic twin peaks, which refer to the points where the s- and r-reaction paths encounter magic neutron numbers. Note that a p process has to be invoked for producing the proton rich nuclei that are not reached by neutron capture reactions. (For details see discussion in text.)
by the solid line in Fig. 2. The s abundances are determined by the respective (n,7) cross sections averaged over the stellar neutron spectrum, such that isotopes with small cross sections are building up large abundances. This holds for nuclei with closed neutron shells giving rise to the sharp s-process maxima in the abundance distribution at A=88, 140, and 208. This represents an illustrative example for the intimate correlation between the relevant nuclear properties and the resulting abundances, a phenomenon that can be used for probing the physical conditions during nucleosynthesis. The r-process counterparts of these maxima are caused by the effect of neutron shell closure on the /3-decay half-lives. Since the r process occurs in regions of extremely high neutron density (presumably during stellar explosions in supernovae) neutron captures are much faster than /3-decays. Therefore, the r-process path is driven off the stability valley until nuclei with neutron separation energies of « 2 MeV are reached. At these points, (n,7) and (7,n) reactions are in equilibrium, and the reaction flow has to wait for /3-decay
5
to the next higher element. Accordingly, the r abundances are proportional to the half-lives of these waiting point nuclei. This means that r-abundance peaks accumulate also at magic neutron numbers, but at significantly lower A compared to the related s-process maxima, resulting in the typical twin peaks of the abundance distribution. While the observed abundances are dominated by the s and r components, which both account for approximately 50% of the abundances in the mass region A>60, the rare proton-rich nuclei can not be produced by neutron capture reactions. This minor part of the abundance distribution had to be ascribed to the p process that is assumed to occur in explosively burning outer shells of supernovae 20>12. Among these processes, the s process is best accessible to laboratory experiments as well as to stellar models and astronomical observations 9 . Attempts to describe the r and p processes are hampered by the large uncertainties in the nuclear physics data far from stability 8 ' 2 1 , but also - and perhaps more severely - by the problems related to a detailed modelling of the stellar explosion 20>22>23. Obviously most isotopes received abundance contributions from the s and r processes. But as indicated in Fig. 2 there are neutron-rich stable isotopes (marked r) that are not reached by the s process because of their short-lived neighbors. Consequently, this species is of pure r process origin. In turn, these r-only nuclei terminate the /?-decay chains from the r-process region, making their stable isobars an ensemble of s-only isotopes. The existence of these two subgroups is of vital importance for nucleosynthesis, since their abundances represent important tests for stellar models.
4
The Case of the s Process
The diret impact of neutron reactions for the processes sketched before is illustrated at the example of the s process. The main nuclear physics input for s-process studies are the (n,7) cross sections averaged over the thermal neutron spectra characteristic for the stellar sites of the s process, typically between T 8 ~ 1 and 3 (in units of 108 K). This information is required for all nuclei along the reaction path from Fe to Bi. In addition, the /?-decay rates for unstable isotopes, which act as branching points in the reaction chain, have to be evaluated 2 4 .
6
4-1
Laboratory Neutron Sources
Neutrons in the energy range between 0.3 and 300 keV required for such measurements are produced in several ways: (i) At low-energy particle accelerators, nuclear reactions, such as 7 Li(p,n) 7 Be offer the possibility of tailoring the neutron spectrum exactly to the energy range of interest. This has the advantage of low backgrounds, allowing for comparably short neutron flight paths to compensate limitations in the neutron source strength 9 ' 25 . (ii) Much higher intensities can be achieved at linear accelerators via (7,n) reactions by bombarding heavy metal targets with electron beams of typically 50 MeV. The resulting spectrum contains all energies from thermal to near the initial electron energy. Since the astrophysically relevant energy range corresponds only to a small window in the entire spectrum, background conditions are more complicated and measurements need to be carried out at larger neutron flight paths. In turn, the longer flight paths are advantageous for high resolution measurements which are important in the resonance region. Refs. 26 27 ' are recent examples of astrophysical measurements at such facilities. (iii) Spallation reactions induced by energetic particle beams provide the most prolific sources of fast neutrons. An advanced spallation source suited for neutron time-of-flight (TOF) work is the LANSCE facility at Los Alamos, allowing for measurements on very small samples as well as on radioactive targets 28 - 29 . While the situation at LANSCE is characterized by a comparably short flight of 20 m and a time resolution of 250 ns (similar to what is planned at the SNS in Oak Ridge) the new n_TOF facility at CERN represents a complementary approach aiming at higher resolution (185 m flight path, 7 ns pulse width) 30 ' 31 - 32 . 4-2
Measurement of Neutron Capture Rates
The experimental methods for measuring (11,7) cross sections fall into two groups, TOF techniques and activations. In principle TOF techniques can be applied to all stable nuclei and require a pulsed neutron source for determining the neutron energy via the flight time between neutron production target and capture sample. Capture events are identified by the prompt 7-ray cascade in the product nucleus. The best signature for the identification of neutron capture events is the total energy of the emitted 7-cascade. To use this feature for accurate (11,7) cross section measurements requires a detector that operates as a calorimeter with good energy resolution such as the Karlsruhe in BaF 2 detector 33 . In the 7-spectrum of a perfect calorimeter, all capture events would fall in a line at the neutron binding energy (typically between 5 and 10 MeV), well separated
7
from backgrounds, which are inevitable in neutron experiments. In this way, an efficiency for capture events of 96 to 98% can be obtained, allowing for cross section uncertainties of ±1 %. Similar calorimeters are presently under construction at Los Alamos and at CERN. Activation in a quasi-stellar neutron spectrum provides a completely different approach for the determination of stellar (n,7) rates, but is restricted to those cases, where neutron capture produces an unstable nucleus. This method has superior sensitivity, allowing to use sub-/ig samples, and is highly selective, which means that isotopically enriched samples are not required. Quasi-stellar neutron spectra can be produced via the 7 Li(p,n) 7 Be 34 ' 35 by bombarding thick metallic lithium targets with protons of 1912 keV, only 31 keV above the reaction threshold. The resulting neutrons exhibit a continuous energy distribution very similar to a Maxwell-Boltzmann distribution for kT = 25 keV. The possibility to use minute samples makes the activation technique an attractive tool for investigating unstable nuclei of relevance for s-process branchings 36 . For example, a measurement of the 155 Eu cross section (ti/2=4-96 yr) could be performed with a sample of only 88 ng corresponding to 3.4xl0 1 4 atoms. This aspect is essential for minimizing the sample activity and, hence the radiation hazard, to a manageable level 3T . 4-3
Theoretical Calculations
In spite of the experimental progress, cross section calculations remain indispensable for determining the (n,7) rates of unstable nuclei with high specific 7-activity as well as the (possible) differences between the laboratory values and the actual stellar cross sections, which can be affected by thermally populated nuclear states with low excitation energies. Theoretical reaction rates are particularly important for explosive scenarios, where nuclei far from stability are involved and where experimental data are completely missing 38,39 Another essential issue are weak interaction rates under astrophysical conditions, both for He burning 24 and explosive scenarios 8 . 5 5.1
s Process Models The Canonical s Process
This phenomenological model 9 ' 40 was suggested by the empirical assumptions that temperature and neutron density are constant during the s-process and that a certain fraction G of the observed 56 Fe abundance was irradiated by an exponential distribution of neutron exposures. Then, an analytical expression
8
can be derived to calculate for all involved isotopes the characteristic s-process quantity, i.e. the product of the stellar cross section and the respective s abundance. Apart from the two parameters G and TQ, which are adjusted to fit the abundances of the s-only nuclei, the stellar (n,7) cross sections (a) are the only input data required for determining the overall abundance distribution. This approach includes also the treatment of the particular sprocess branchings. Given the very schematic nature of the classical approach, it was surprising to see that it provides an excellent description of the s-process abundances. Fig. 3 shows the calculated (cr)Ns values compared to the corresponding empirical products of the s-only nuclei (symbols) in the mass region between A = 56 and 209. The error bars of the empirical points reflect the uncertainties of the abundances and of the respective cross sections. One finds that equilibrium in the neutron capture flow was reached between magic neutron numbers, where the (cr)A?s-curve is almost constant. The small cross sections of the neutron magic nuclei around A~88, 140, and 208 act as bottlenecks for the capture flow, resulting in the distinct steps of the crN-curve. 5.2
Stellar Models
In terms of stellar sites, the main component can be attributed to helium shell burning in low mass stars, where neutron production and concordant s-processing occur in two steps by the 1 3 C(a,n) 1 6 0 reaction at relatively low temperatures around T g ~ l and by the 22 Ne(a,n) 25 Mg reaction at Tg~3 (see Refs. 10 ' 41 for details). The weak component can be ascribed to core He burning in massive stars 42 . 6
s-Process Branchings
Branchings in the reaction chain of the s process occur at unstable nuclei with sufficiently long half-lives that neutron capture can compete with /?decay. The resulting abundance pattern provide direct clues with respect to stellar neutron density, temperature, and pressure and allow to characterize the He-burning zones, where the s process actually takes place. Fig. 4 shows the s-process branchings at 147 Nd and 1 4 7 ' 1 4 8 Pm, which are defined by the sonly nuclei 148 Sm and 150 Sm. Since 148 Sm is partly bypassed by the reaction flow, its (<J)NS value will be smaller than that of 150 Sm, the ratio providing a measure for the combined strength of the branchings. Quantitative branching analyses require (i) the cross sections of the involved s-only nuclei with uncertainties of « 1%, and (ii) the corresponding
9
g 1000
z < Q
z n < z o p o PJ o ft! U
100
10 r
0.1
100
150
200
MASS NUMBER Figure 3. The characteristic product of cross section times s-process abundance plotted as a function of mass number. The solid line was obtained via the classical model, and the symbols denote the empirical products for the s-only nuclei. A complete representation of the empirical values requires at least two different mechanisms, the m a i n and weak (thick and thin solid lines, respectively). The important branchings of the neutron capture chain are indicated as well.
cross sections of the radioactive branch point isotopes with uncertainties of « 5 to 10%,. At present, the lack of experimental information on unstable isotopes is the limiting problem for reliable branching, because statistical model calculations are bound to uncertainties of 20% to 30%. Therefore, future experimental efforts have to be directed to determine cross sections of unstable nuclei 36 ' 43 ' 44 . Since the /3-decay rates of the branch points at A= 147-149 in Fig. 4 are not significantly affected by temperature 24 , these branchings are suited for determining the neutron density. A measurement with the 47r BaF2 detector yields fg = 0.870 ± 0.009 45 leading to an effective neutron density of (4.1 ± 0.6) • 108 cm" 3 4 6 . 7
N e u t r o n D a t a for A s t r o p h y s i c s : S t a t u s a n d N e e d s
The present status of (11,7) data for the s process is summarized in the compilation of Bao et al. 47 . In short, it can be stated that experimental techniques
10 p process
M7
H8
Sra
Sm
\ m
150
""Sm
148pm
Pm
Sm
k
\ 149
Pm
146Nd
\
\
\ —
—
147
Nd
148
Nd
150 N( J
5 process
r process Figure 4. The s-process reaction path in the N d / P m / S m region with the branchings at A=147, 148, and 149. Note that 1 4 8 Sm and 1 5 0 Sm are shielded against the r process. These two isotopes define the strength of the branching.
have reached a stage where the 1% accuracy level required for meaningful analyses of particular abundance patterns can been met, but that this has been achieved so far only for a minority of the relevant isotopes. Apart from the remaining key isotopes, also a large number of cross sections with uncertainties in excess of 10% await improvement. In contrast to the comparably stable situation of the s process, the complex explosive nucleosynthesis scenarios imply huge reaction networks including several thousand reactions. Since the majority of the reaction rates has to be obtained by statistical model calculations, experimental data for stable and as many unstable isotopes as possible are, therefore, required to test the necessary extrapolation to the unstable nuclei of relevance for these networks. In the following, the principal data needs for quantitative nucleosynthesis studies in the heavy element region are schematically summarized. For s-process analyses the requests concentrate on (n/y) measurements in the following areas: • The s-only nuclei are the key isotopes for all s-process investigations including the analyses of s-process branchings. These cross sections should, therefore, be determined with uncertainties of RJ 1%. So far, this has been reached only for half of the 33 s-only nuclei between T0 Ge and 2 0 4 Pb.
11 • Meaningful analyses of the characteristic signatures preserved in presolar grains require also accurate cross sections with uncertainties of ft* 1%. However, the present status is far from being adequate, particularly for the lighter elements oxygen, neon, magnesium, silicon, calcium, titanium, and zirconium. In this group, d a t a for about 70 isotopes have to be determined. • Nuclei at or near magic neutron numbers N = 5 0 , 82, and 126, which act as bottlenecks for the reaction flow in the main s-process region between Fe and Bi. For the majority of these d a t a the necessary uncertainties of < 3 % have not been reached. • T h e cross sections of a b u n d a n t light isotopes below Fe, which may constitute crucial neutron poisons for the s-process, need to be improved. Of particular importance are 1 6 0 , 1 8 0 , and 2 2 Ne. • Cases where Direct C a p t u r e (DC) contributes significantly to the astrophysical reaction rate are of particular interest, because this effect plays an important role in neutron-rich nuclei. Interesting examples are 2 0 8 P b , 14 C , 1 6 0 , 8 8 Sr, and 1 3 8 B a . • Nuclei, which still constitute white spots in the s-process chain or which exhibit very uncertain cross section, are found in the mass region below Fe, around A = 1 0 0 , and near the end of the s-process region. These gaps in the experimental d a t a should be determined at the 5% level. • Last, but not least, enhanced efforts should be directed to measurements on unstable nuclei. In addition to the activation technique, the very high neutron fluxes available at spallation neutron sources appear to be promising options for such studies 48 > 49 . From a list of possible measurements, priority should be given t o the important branch points 7 9 Se, 147 P m , 1 5 1 Sm, 1 8 3 H o , 1 7 0 T m , 1 7 1 T m , 1 7 9 Ta, 2 0 4 T 1 , and 2 0 5 P b . These cases are of immediate relevance to s-process analyses and should not present unexpected experimental problems. In addition to this list, partial cross sections leading to long-lived isomers are important for several branchings. A well-known example is the 10.8 yr isomer in 8 5 K r , which determines the s abundance of the neutron magic isotope 8 6 K r . While such studies were previously limited to activation measurements at very few energies, recently T O F measurements of partial cross section with a total absorption calorimeter have been reported 5 0 , yielding the energy-dependence of the partial cross sections, which is necessary to follow
12 the evolving abundance patterns during the complex He burning scenarios. Similarly, elastic and inelastic scattering data are definitely needed for establishing a quantitative set of stellar enhancement factors, in analogy to the treatment of the Os isotopes 51 ' 52 . Finally, the neutron producing (a,n) reactions on 13 C and 22 Ne exhibit large uncertainties and are not yet directly measured in the stellar energy range. The extrapolation of existing data to stellar energies requires a comprehensive R-matrix analysis, which combines all relevant reaction channels. Accordingly, measurements of the (n,a)-cross sections of 1 6 0 and 25 Mg would provide a significant contribution. Since neutron data for explosive nucleosynthesis are completely missing, any effort in this area provides a most useful support for testing and amending the theoretically calculated rates, which are used in the network calculations. In the r-process, neutron cross sections have a direct impact for scenarios with comparably low neutron densities as well as during freeze-out, where they contribute to smooth the pronounced odd-even effects predicted for the primary yields. In principle, several unstable nuclei on the neutron-rich side of the stability valley could be studied experimentally, e.g. 90 Sr, 123 ' 126 Sn, 182 Hf, 228 Ra, and a number of higher actinides. Such data would also improve the description of freeze-out effects in the p process, where neutrons are liberated by (7,11) reactions during the explosive burning of the Ne/O shell 53 . Furthermore, (n/y) cross sections of protonrich nuclei would be most useful in determining the inverse rates by detailed balance. Experimentally feasible cases include about 25 unstable isotopes between 53 Mn and 2 0 2 Pb. Apart from measurements on unstable nuclei, even data for stable isotopes are urgently required for improving the reaction rates used in explosive nucleosynthesis. In particular, complete data sets for long isotope chains are important for this purpose. This means that stellar (11,7) cross sections should be determined also for all r- and p-only nuclei. At this point it should be mentioned that there are only very few experimental (p,7) and (0,7) cross sections at astrophysically relevant energies in the mass region of the p process, even for stable isotopes. Especially, the (a/y) and (a,p) rates lead to significant uncertainties in the final p-process abundances. Since direct measurements are difficult and time-consuming, the calculated rates are poorly constrained by experimental data. In particular, the a-nucleus potential used in statistical model calculations seems to be rather uncertain. A series of (n,a) measurements at astrophysically meaningful energies could help to solve this persisting problem.
13 8
Summary
Neutron capture nucleosynthesis of the elements heavier than iron operates during the He burning stages of stellar evolution and (presumably) in the final supernova explosion of massive stars. The various scenarios are identified by their typical abundance distributions as well as by increasingly detailed astronomical observations. This information combined with quantitative model calculations allow to probe stellar and galactic evolution. In this context, the strong demand for reliable nuclear physics data represents a continuing challenge, requiring the implementation of more powerful neutron sources and new experimental techniques. References 1. V.M. Goldschmidt, Norske Vidensk. Akad. Skr., Mat.-Naturv. Kl. IV, 1937. 2. H.A. Bethe and C. Critchfield, Phys. Rev. 54, 248 (1938). 3. C.F. von Weizsacker, Physik. Zeitschrift 39, 639 (1938). 4. H.A. Bethe, Phys. Rev. 55, 103 (1939). 5. P.W. Merrill, Science 115, 484 (1952). 6. E.M. Burbidge, G.R. Burbidge, W.A. Fowler, and F. Hoyle, Rev. Mod. Phys. 29, 547 (1957). 7. G. Wallerstein et al., Rev. Mod. Phys. 69, 995 (1997). 8. F. Kappeler, F.-K.. Thielemann, and M. Wiescher, Ann. Rev. Nucl. Part. Sci. 48, 175 (1998). 9. F. Kappeler, Prog. Nucl. Part. Phys. 43, 419 (1999). 10. M. Busso, R. Gallino, and G.J. Wasserburg, Ann. Rev. Astron. Astrophys. 37, 239 (1999). 11. T.J. Bernatowitz and E. Zinner, eds., Astrophysical Implications of the Laboratory Study of Presolar Material, (AIP, New York, 1997). 12. D.L. Lambert, Astron. Astrophys. Rev. 3, 201 (1992). 13. J.J. Cowan, F.-K. Thielemann, and J.W. Truran, Phys. Rep. 208, 267 (1991). 14. E. Zinner, Ann. Rev. Earth Planet. Sci. 26, 147 (1998). 15. H. Schatz et al., Physics Reports 294, 167 (1998). 16. E. Anders and N. Grevesse, Geochim. Cosmochim. Acta 53, 197 (1989). 17. H. Palme and H. Beer, in Landolt-Bornstein New Series, Group VI, Vol. VI/3a, Astronomy and Astrophysics, ed. O. Madelung (Springer, Berlin, 1993), page 196. 18. C. Sneden, J.J. Cowan, D.L. Burris, and J.W. Truran, Ap. J. 496, 235
14
(1998). 19. C. Sneden, Nature 409, 673 (2001). 20. M. Rayet et al., Astron. Astrophys. 298, 517 (1995). 21. F.-K. Thielemann et aJ., in Nuclear and Particle Astrophysics, eds. J.G. Hirsch and D. Page (Cambridge University Press, Cambridge, 1998), page 27. 22. W. Hillebrandt and P. Hoflich, Rep. Prog. Phys. 52, 1421 (1989). 23. E. Miiller, in Nuclear Astrophysics, eds. M. Buballa, W. Norenberg, A. Wambach, and J. Wirzba (GSI, Darmstadt, 1998), page 153. 24. K. Takahashi and K. Yokoi, Atomic Data Nucl. Data Tables 36, 375 (1987). 25. Y. Nagai et al, Ap. J. 381, 444 (1991). 26. P.E. Koehler et al, Phys. Rev. C 54, 1463 (1996). 27. H. Beer, F. Corvi, and P. Mutti, Ap. J. 474, 843 (1997). 28. P.E. Koehler and F. Kappeler, in Nuclear Data for Science and Technology, ed. J.K. Dickens (ANS, La Grange Park, Illinois, 1994), p. 179. 29. R.S. Rundberg et al, Technical report, Los Alamos National Laboratory, Los Alamos, USA (1999). 30. C. Rubbia, et al. Technical report, CERN, Geneva, Switzerland (1998). 31. S. Andriamonje et al, Report CERN/INTC 2000-004, CERN, Geneva, Switzerland (2000). 32. U. Abbondanno et al, Report CERN/INTC 2001-021, CERN, Geneva, Switzerland (2001). 33. K. Wisshak et al, Nucl. Instr. Meth. A 292, 595 (1990). 34. H. Beer and F. Kappeler, Phys. Rev. C 2 1 , 534 (1980. 35. W. Ratynski and F. Kappeler, Phys. Rev. C37, 595 (1988). 36. F. Kappeler, M. Wiescher, and P.E. Koehler, in The Production and Use of Intense Radioactive Beams at the Isospin Laboratory, ed. J.D. Garrett (Joint Institute for Heavy Ion Research, Oak Ridge, 1992), p.163. 37. S. Jaag and F. Kappeler, Phys. Rev. C 5 1 , 3465 (1995). 38. T. Rauscher and F.-K. Thielemann, Atomic Data Nucl. Data Tables 75, 1 (2000). 39. J. Goriely, Long Term Needs for Nuclear Data Development, ed. M. Herman (International Atomic Energy Agency, Vienna, 2001), p.83. 40. P.A. Seeger, W.A. Fowler, and D.D. Clayton, Ap. J. Suppl. 97, 121 (1965). 41. R. Gallino et al, Ap. J. 497, 388 (1998). 42. C M . Raiteri et al, Ap. J. 419, 207 (1993). 43. S. Jaag, F. Kappeler, and P.E. Koehler, Nucl. Phys. A 621, 247c (1997). 44. J.B. Wilhelmy et al, in International Conference on Nuclear Data for
15
Science and Technology, Tsukuba, Japan, October 7 to 12 (2001). 45. K. Wisshak et al., Phys. Rev. C 4 8 , 1401 (1993). 46. F. Kappeler, K.A. Toukan, M. Schumann, and A. Mengoni, Phys. Rev. C 5 3 , 1397(1996). 47. Z.Y. Bao et al., Atomic Data Nucl. Data Tables 76, 70 (2000). 48. A.F. Michaudon and S.A. Wender, Report LA-UR-90-4355, Los Alamos National Laboratory, Los Alamos, USA (1990). 49. S. Abramovich et al, Report CERN/SPSC 99-8; SPSC/P 310, CERN, Geneva, Switzerland (1999). 50. K. Wisshak, F. Voss, C. Arlandini, F. Kappeler, and L. Kazakov, Phys. Rev. C 6 1 , 065801 (2000). 51. R.R. Winters, R.F. Carlton, J.A. Harvey, and N.W. Hill, Phys. Rev. C 34, 840 (1986). 52. R.R. Winters, R.L. Macklin, and R.L. Hershberger, Astron. Astrophys. 171, 9 (1987). 53. M. Rayet, N. Prantzos, and M. Arnould, Astron. Astrophys. 227, 271 (1990).
THE DETECTOR FOR ADVANCED NEUTRON CAPTURE EXPERIMENTS AT LANSCE J.L. ULLMANN, R.C. HAIGHT, L. HUNT, E. SEABURY, R.S. RUNDBERG, J.B. WILHELMY, M.M. FOWLER. D.D. STROTTMAN Los Alamos National Laboratory, Los Alamos NM87544,
USA
F. KAEPPELER, R. RE1FARTH, M. HEIL Forschungszentrum
Karlsruhe, Karlsruhe,
Germany
E.P. CHAMBERLIN Chamberlin Associates, Los Alamos, NM87544
USA
The Detector for Advanced Neutron Capture Experiments (DANCE) is a 159-element 4tt barium fluoride array designed to study neutron capture on small quantities of radioactive material. It is being built on a 20m neutron flight path which views the "upper tier" water moderator at the Manuel J. Lujan Jr. Neutron Scattering Center at the Los Alamos Neutron Science Center. Monte Carlo calculations have suggested ways to minimize backgrounds due to neutron scattering events. Preliminary data on an 8 mg sample of 234U and a 0.5 mg sample of l!l Sm have been taken using dUf, detectors.
1
Introduction
The precise measurement of neutron capture cross sections in the electron-volt and kilo-electron-volt regions on radioactive isotope targets is needed for several applications, including stockpile stewardship and nuclear astrophysics. Capture cross sections are difficult to calculate accurately because they depend on fine details of nuclear structure and level densities at 5 to 10 MeV in excitation. A recent compilation of "Maxwell-averaged" capture cross sections using the "NonSmoker" statistical model code [1] showed that the calculated cross sections differed from measured cross sections by +/- a factor of 2 for masses between 25 and 210. While there are capture measurements on most stable nuclides, there are very few measurements on unstable nuclides One of the main applications for capture cross sections is in understanding sprocess nucleosynthesis [2]. The s process occurs by sequential neutron capture along the line of beta stability. When the capture sequence produces an unstable nuclide, the process can branch. The competition between beta decay and neutron capture at the branch nuclide depends on it's stellar beta-decay half life, the stellar neutron density, and the capture cross section. Combined with observed nuclear abundances, the capture cross section and beta-decay rate can be used to infer the temperature and neutron density at the stellar s-process site. The interesting energy range (Fig. 1) for these cross sections is over a Maxwell distribution centered at 25
16
17 keV, for neutrons from the 22Ne(a,n) reaction, and at about 10 keV, for the 13C(oc,n) reaction. Maxwell Distribution 12.0 10.0 8.0
30keV
,-\
- - - -10keV 1 /
6.0 4.0
!/
2.0 0.0
, ' • • -
•.• —
100
50
200
150
Energy (keV)
Figure 1. Maxwell energy distribution
A second application is in Stewardship Science, where capture cross sections on unstable nuclides are needed to interpret "rad-chem" diagnostics. Stable isotopes were placed in past nuclear explosion tests as diagnostic aids which integrated the neutron exposure over the entire explosion history. The high neutron densities caused multiple reactions creating many nuclei (See Fig 2.) Neutron cross sections are needed for neutron energies up to about 1 MeV.
(n,2n) !68Tm
4
4
"'"Tm •
93 d
(n,2n)
(11,7)
• Stable (11,7)
(n,2n) ,70
4
(n,2n) 171
Tm
4
Tm
• 129 d
(11,7) 1.9 y
172
• OVy)
2J
Tm d
Figure 2. An example of a reaction sequence where cross sections needed for "rad-chem" diagnostics
Three separate experimental components are needed to make these measurements: An intense neutron source, facilities to fabricate and handle radioactive targets, and an efficient, well characterized gamma detector. The source and detector are discussed further below; target preparation is the subject of a separate talk at this conference [3].
18 2
Neutron Source
The DANCE is being constructed on Flight Path 14 at the Manuel J. Lujan, Jr. Neutron Scattering Center at LANSCE. Flight Path 14 views the upper-tier "backscatter" water moderator, from which there are significantly fewer neutrons above 1 MeV. The sample is positioned 20 m from the moderator and the beam stop is at 30 m. A box for remotely inserting various filters is at 7 m. The bulk shielding surrounding the spallation target is 4.72 m in radius. Four discrete collimators are located outside the bulk shield, each constructed of about a meter of copper, brass, and 5% borated polyethylene. The collimation was designed to produce a uniform 1 cm diameter beam spot at the target location with minimal penumbra outside the central beam. The last collimator has a r = 0.3 cm opening with the downstream edge at 18.88 m. This tight collimation limits the beam intensity, which depends on the area of the moderator that is viewed. To reduce gamma backgrounds, the beam pipes and flanges were constructed of aluminum and the use of iron in components outside of the bulk shield was quite to a minimum. The flight path shielding was designed to limit the total gamma plus neutron dose to less than 1.0 mrem/hr along the first 10 m of the flight path and 0.5 mrem/hr beyond 10 m. Magnetite-loaded concrete blocks were used to shield the beam pipes and target area. Only polyethylene and borated polyethylene were used for shielding the roof of the target area. The Monte Carlo shielding calculations predicted significant high-energy gamma-ray production from neutron capture in the polyethylene and concrete, and the interior walls of the target area were faced with 2.54 cm thick 5% borated polyethylene which yields lower energy gamma rays following neutron absorption.. During the 2001 run cycle, the spallation source was operated at 55 uA. The neutron flux on FP14 was measured with three different techniques. First, a standard 3He tube was used [4]. Next, a fission chamber with 286 u,g/cm2 of 235U was used. Lastly, a neutron monitor consisting of a 546 u,g/cm2, 1 cm diameter deposit of 6LiF on an Al foil backing and viewed by a Si surface barrier detector, was employed to detect neutrons via the 6Li(n,at) reaction. The measured flux is shown in Fig. 3. The three measurements were each made at a different location, and were converted to moderator surface current for comparison. The three measurements were not consistent, and were considerably below the anticipated flux. This discrepancy is not understood and is still under study. It may possibly be due to misalignment. The 3He and 6Li measurements can be fit to a surface current of the form / = A/E with E in eV and A = 1.10 x 10 N/cm2/sr/eV/sec at 55 uA. At 20 m, this yields a flux of O = (3.70 x 103 N/cm2/eV/sec)/E.
19 FP-14 Flux 1.E+10 •
1.E+09
u
1.E+08 1.E+07
"
\
1 * n — •
r.
:
_
!
He-3 Geo Li6Sum265 55 uA U235Run128
-
: ^^D»
1.E+06
: 0111
1.E+05 1.E+04
«£#*> %
:
i%
•
;
1.E+03 1.E-01
in
i 1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05 1.E+06
Energy (eV) Figure 3. Measured surface currents compared to expected value.
3
Preliminary Data with C6D6 Scintillators
Preliminary data was taken using two C6D6 detectors, each 12.5 cm diameter by 7.5 cm thick, mounted adjacent to a 5 cm diameter target holder made from 0.16 cm Al tubing. The targets studied were Au, ""•"°'^u powder, each contained in a small quartz tube, l50,l52Sm powder in quartz, and 151Sm deposited on a thin Ti foil. These data are still being analyzed. Figure 4 shows the time-of-flight spectrum from 234U target, which was 8 mg of 99.8% pure material prepared by the Chemistry Division. The time scale is 100 ns/chan, and the large resonance near channel 6500 is at 5.16 eV. Figure 5 shows the time-of-flight spectrum for the 0.5 mg 'Sm target. The isotopic composition of this target is 71% 151Sm, 13% 150Sm, 7% 152Sm, 6% l47Sm, and 1% ,49Sm. The large resonance near channel 14000 is the 1.088 eV resonance in l5, Sm. Note that 1 keV corresponds to channel 500. The gray line is the spectrum from a Ti foil blank normalized to the same neutron fluence. 4
DANCE Design
The advanced gamma detector is being built to provide an increased and betterdetermined detection efficiency and also to provide better background rejection. Backgrounds are due to capture in material surrounding the target, but also due to
20
Figure 4. Time of flight spectrum for 234U(n,y). The horizontal scale is 100 ns/channel.
I Sum GEOATI
1000
etn aoa 400
200-
Qt—• • • I 2000
4000
8000
6000
10000
12O0Q 14Q0Q
19000
Figure 5. Time of flight spectrum for lslSm(n,Y). The contribution from the blank Ti foil is also shown.
capture reactions in the detector from neutrons scattered in the target. The neutron scattering cross section is greater than the capture cross section for many materials in the kilovolt region. Three criteria were established for the detector: •
Calorimetric to measure the total gamma ray energy emitted
21
• •
Insensitive to neutrons Segmented and fast to handle radioactive targets (one Curie is 37 decays/ns)
Extensive Monte Carlo calculations were made using GEANT-3 to design the detector [5][6]. Of the commonly available scintillator materials, BaF2 was chosen because it had the smallest neutron capture cross section and a very fast (0.6 ns) component of light. It suffers from an internal alpha particle background due to decay of Ra and its decay chain products, but pulse-shape discrimination can be used if needed. The crystal array should completely cover 4% sr with no gaps, and each crystal should have equal area and volume. The analysis of Habs [7] showed that 162 elements with 4 different shapes will meet this requirement. The array is shown schematically in Fig. 6. The inner radius is 17 cm and each crystal is 15 cm deep, 734 cm3 in volume, and has an inside area of 22.9 cm2. Each crystal is coupled to an Electron Tubes 9921 7.5 cm phototube with quartz window using Dow-Corning Sylgaard 184 for maximum UV transmission.
Figure 6. Schematic design of the DANCE crystal ball. Four different shaped crystals are needed, indicated by different shadings in the figure.
22
The Monte Carlo calculations indicated that scattered neutrons could still contribute a significant background, especially in the 10 to 100 keV range of interest. Several additional methods will be employed to decrease this background. First, the measured reaction Q value can be used in many cases to discriminate between true capture events and events induced by scattered neutrons [5,6]. Next, the Monte Carlo calculations predict that an 8 cm thick 6LiH shell inside the array will reduce the scattered neutrons to 42% of the unattenuated number, in the 10 to 100 keV energy range[6]. However, because of space limitations we will use a 6 cm thick 6LiH sphere surrounding the target. Finally, a "hit pattern" analysis of the event will also be tried. This is illustrated in Fig 7. which shows the calculated number of crystal clusters for events due to neutron scatter and true capture in the target [6]. A cluster is a set of adjacent crystals that give a signal above threshold. True capture events produce several clusters, each due to an individual gamma ray from the decay cascade, while events due to neutron capture in the BaF2 tend to be grouped primarily into one cluster. Neutron Energy = 10 to 100 keV
- - - - Scattered
3500
Capture
3000 2500 2000 1500
1
1000 500 n 0
1
2
3
4
5
6
7
Cluster Multiplicity
Figure 7. Cluster multiplicity (see text) for events due to true capture in the target and capture of neutrons scattered into the BaF2 array.
The data acquisition system will consist of two Acqiris DC-265 8-bit waveform digitizers on the anode signal of each phototube. The digitizers sample at 500 MHz. Each digitizer has a different voltage gain to match the dynamic range of the fast and slow components of the scintillation light. The slow component contains about 85% of the light, and has a decay time of 630 ns. The fast component, while providing only 15% of the light, has a decay time of 0.6 ns and is the dominant feature of the waveform. The front end software will initially return for each event only a time and pulse height number, which the analyzer routine will histogram and log.
23
Figure 8 shows a typical pulse from a completed crystal assembly, acquired with an Acqiris DC-270 1 GHz digitizer. Fig 9. shows a 60Co spectrum obtained by simply adding the counts in a waveform from 10 ns before the trigger time to 1790 ns after. The resolution of the 1173 keV peak is 9.0% fwhm. Waveform
^
^
t n tries Mean RMS
^
D~ 508.1 57S.6
'100 80 60 40 20
-500
500
1000
1500
2000 Channel
Figure 8. Typical waveform from a completed BaF2 crystal assembly using a DC-720 1 GHz digitizer. The fast and slow components are easily recognized. Fast + Slow |
sum Entries 329807 Mean 2.173e+04 RMS 79SG
1600 1400
fl
1200
I
~r
/V
800
r
/ 1/L^
600
r
1000
400
rl
I
VI
-
\ r\A
200
M 10000
,
A 20000
,
i 30000
^""V*. 40000
50000
Figure 9.60Co spectrum obtained by simply summing waveforms from a 1GHz digitizer. The peaks from the alpha-decay background in the crystal are readily seen.
24
5
Summary and Future Plans
The DANCE array is under construction, and all crystals are scheduled to be delivered by Sept, 2002. A multi-year program of targets to be measured for stockpile stewardship and s-process branch point studies has been mapped out. Initially, targets that can be chemically purified have been chosen, and 146Nd, 154Sm, and 170Er will be irradiated at the ILL in spring, 2002, to produce 4 to 10 mg of 17 Pm, 155Eu, and m Tm. It is expected that these isotopes will be studied using the DANCE array in 2002 , along with a new measurement of lsl Sm. References 1. T. Rauscher and F.-K. Thieleman. in Atomic and Nuclear Astrophysics, A. Mezzacappa, ed, IOP, Bristol, (1998), p. 519. 2. F. Kaeppeler. Prog. Part. Nucl. Phys. 43, 419-483 (1999). 3. R.S. Rundberg et al., these proceedings. 4. L. Daemen, private communication. 5. M. Heil, et al., Nucl. Instr. Meth. A459, 229-246 (2001). M. Heil, et al, Los Alamos National Laboratory report LA-UR-99-4046, (1999). 6. R. Reifarth et al., Los Alamos National Laboratory Report LA-UR-014185,(2002). 7. D. Habs, F.S. Stephens, and R.M. Diamond, Lawrence Berkeley Laboratory Report LBL-8945, (1979).
A S T R O P H Y S I C S P R O G R A M AT T H E C E R N n_TOF FACILITY A. MENGONI CERN, EP Division, CH-1211 Genava 23, Switzerland e-mail: alberto.mengoniQcem.ch and The n-TOF Collaboration The set of measurements of neutron capture cross sections for nuclear astrophysics at the CERN neutron time-of-flight facility, n_TOF, is presented. A brief description of each of the planned measurements is given. 1
Introduction
The CERN neutron time-of-flight facility1, n_TOF, is a spallation neutron source based on the high-intensity 20 GeV proton beam of the CERN PS accelerator complex. At n_TOF, a solid 80x80x40 cm 3 lead target coupled to a 5cm-thick water moderator generates a white neutron spectrum of extremely high intensity. The proton beam intensity is typically 7 x 1012 protons/pulse, with a pulse-width of 7 ns. Each proton generates 360 neutrons in the spallation process, a fraction of which can leave the target-moderator assembly to enter the 200m neutron flight-path. The first experimental area (EAR-1) is located at 187.5 meters downstream from the lead target. The neutron energies effectively considered for cross section measurements is in the range 1 eV up to 250 MeV, although neutrons from thermal up to the GeV region are generated by the target-moderator assembly. The neutron flux in EAR-1 can reach 4 x 105 neutrons/cm 2 /pulse. The long flight-path, combined with the time-synchronization characteristics of the lead slowingdown process in the spallation target results in an excellent energy resolution, which reaches 3 x 10~ 4 at En « 3 eV and 1.5 x 10" 3 at En « 30 keV. A peculiar feature of n_TOF is the very low repetition frequency of the PS beam. On average, 1 pulse/2.4 s is extracted and delivered to the n_TOF experimental area. A very low ambient background has been recently achieved in EAR-1 after the implementation of meter-thick concrete and iron walls in the tof tunnel. The prompt flash in EAR-1 due to minimum ionizing particles has been drastically reduced by this shielding. At the same time, the fluence of negative muons, responsible for neutron capture background generation in the experimental area, has been largely reduced 2 , such that sample-induced
25
26 backgrounds are the dominant component in E A R - 1 , thus defining optimal conditions for capture measurements. n . T O F has been constructed with t h e basic motivations of measuring: • cross section relevant to nuclear waste transmutation and related nuclear technologies, • neutron cross sections relevant for nuclear astrophysics, and • cross sections for nuclear structure studies. Here we will describe the experimental program for the measurements of interest in nuclear astrophysics. 2
P r i o r i t y m e a s u r e m e n t s for n u c l e a r a s t r o p h y s i c s
n_TOF delivers neutrons in a wide energy range (from eV t o MeV), covering entirely the maxwellian velocity distributions corresponding to a wide range of stellar temperatures, typical for neutron capture nucleosynthesis. In particular, the extended range of t e m p e r a t u r e s recently considered in stellar models, kT « 5 keV up to kT « 30 keV, can be all explored in neutron c a p t u r e measurements for s-process nucleosynthesis. This, combined with the high neutron flux, led to consider t h e possibility of measuring of small-mass samples, either rare o r / a n d radioactive. In addition to t h e high-flux, another favorable condition for measurements of capture cross sections of radioactive samples has been taken into consideration: the low repetition frequency of t h e neutron pulses. To give a quantitative estimate of this condition, one can compare the repetition frequency, / , and flight-path, L, at which two facilities, n_TOF and GELINA, have approximately the same flux at 25 keV (respectively 187m for n J T O F and 50m for GELINA). One finds /i_ /2
Lx X
L2~
0.28 s " 1 800s-
1
187.5 m
1
50m
762'
Obviously, factors of this magnitude will introduce a considerable advantage in t h e performance for c a p t u r e measurements on radioactive samples, often required in t h e studies of s-process branchings. Another characteristics which represents an advantage in capture measurements at neutron spallation sources such as n_TOF over electron-based installations, is t h e large suppression of t h e 7-fiash due to a reduced 7-ray production in the spallation process.
27 Table 1. Priority measurements for nuclear astrophysics at n_TOF.
Reaction
Notes
151
s-process branching(s) at A w 150. tive (t 1 / 2 = 93 yr)
Sm(n, 7 ) 1 5 2 Sm
186 187 188
'
-
Os(n, 7 )
151
Sm is radioac-
nuclear cosmochronology (Re/Os clock)
24 25 26
isotopic abundance ratios in interstellar grains. Relative importance of the 22 Ne(a, n) 25 Mg as neutron source for the s-process. Light nuclei, small cross sections
204 206 208
termination of the s-process. Small cr7/o"e;
- < Mg(n, 7 )
209
- - Pb(n,7), Bi(n, 7 )
Given these preliminary considerations, the priority list shown in Table 1 for the first period of measurements has been defined. We will give here a brief description of the motivations and of the technique to be used in each of the neutron capture cross section measurements listed in Table 1. 2.1
151
Sm(n,7) 1 5 2 Sm
The s-process reaction flow in the Sm-Eu-Gd region exhibit several branching points. These are due to a sufficiently long /3-decay half-life of nuclei which can compete with the long half-life for neutron capture in s-process-like stellar conditions. One specific and crucial example is 151 Sm. Its neutron capture cross section is expected to be quite large, but experimental values in the keV energy region are completely missing. We plan to measure the 151 Sm(n, 7) 152 Sm cross section at n_TOF with a 180 mg sample provided by Oak Ridge National Laboratory. The experimental setup for this measurement includes a set of two CeD 6 -based liquid scintillator detectors, specifically designed with carbon-fiber container, to reduce their sensitivity to scattered neutrons. The high instantaneous n_TOF neutron flux requires the use of a specifically designed data acquisition system, entirely based on high-frequency flash-ADC, operating at 1 GHz sampling rate with 16 MB buffer memory per channel.
28
Given the expected high o-y/aei for the neutron induced reactions on Sm, favorable background conditions are expected for this capture measurement. The data taking is expected to start in June 2002. 151
2.2
186 187 188
'
'
Os(n ) 7 )
The enhancement of the observed 187 Os solar abundance, as compared with what can be expected from s-process synthesis in the A « 185 mass region, is to be attributed to the 187Re(/3~) —• 187 Os decay. 187 Re has a half-life of 42.3 ± 1.3 x 109 yr and it is an r-only isotope. On the other hand, 1 8 6 0s and 187 Os are shielded against r-process production by 1 8 6 W and 187 Re: they are s-only isotopes. A simple calculation based on the canonical s-process, shows immediately the necessity for the radiogenic production of 187 Os. Prom the s-process condition Cyi[A] = const, it follows that a 186 [ 186 Os] = <x187[187Os] From solar-system abundances, rl87 Os] [186 Os]
0.79.
On the other hand, p
= ( £i56 I CT
187/
=
0.48
exp.
using the experimental capture cross sections of Winters et al.3. From the amount of radiogenic 1 8 7 0s, it is possible to derive, within this simple picture, the galactic age. It is also possible to show that the uncertainty in the capture cross section ratios propagates into an age (to) uncertainty of the order of ARa « 10%
==>
Ai 0 « 3 Gyr
The present estimated uncertainty in the capture cross section ratios, Ra, is of the order of 5%. However, the need for maxwellian averages at low kT (kT ss 8 keV) calls for an extension of the measured data to a wider neutron energy range, in particular on the low energy side (see Figure 1). Clarification of other nuclear physics aspects related to this clock will also benefit from measurements of capture cross sections of the other Os isotopes. A comprehensive analysis of the s-processing based on low-mass AGB stars, together with up-to-date galactic chemical evolution models are part of the n_TOF activities on this subject.
29
Os(n,y)
Os
Cross Section and Maxwellian Average (MACS) I
I I I
" " 1
'
'•
1
i
i i 11111
i
i
— HF calculation • Exp.
=
10
^X5£-
1 --
0.1 0.0001 1 —
-— —
O 0.6 0.4
-
- ^*\»
1
1
1
I 0.01
1
0.001 1
1
1
i
r
0.1
1 1 1 1 1 1
kT = 30KeV kT - 8 KeV
-
-
-
-
-
-
0.2
0 0.0001 Wed Oct 16 17:50:43 2000
0.001
0.01
0.1
Neutron energy [MeV]
Figure 1. The neutron capture cross section of 1 8 6 Os (upper panel). A comparison between the experimental values of Winters et al. 3 and a calculation based on Huser-Feshbach statistical model theory, is shown. In the lower panel, the cumulative maxwellian average, vs neutron energy, assumed as upper-limit in the averaging integral, is shown for two values of the stellar temperature kT.
2.3
24 25 26
> - Mg(n, 7 )
The absence of a 25 Mg anomaly in interstellar grains and the role of the 22 Ne(a, n) 25 Mg in AGB stars represent the basic motivations for the 24 25,26 ' Mg(n, 7) measurements. In addition, nuclear structure properties as well as reaction mechanisms aspects, such as the contribution of a direct capture process in the keV neutron energy region, can be derived from these measurements. From the experimental point of view, this set of measurements is challenging, as for these low-mass targets, the ay/aei is very unfavorable. Here, the problem is represented by the scattered neutrons. These generate additional background signals in the experimental area. To partly cure this problem, a
30
set of specifically designed, low neutron-sensitive CeD6 detectors have been constructed and will be used in these measurement.
2.4
204
>206-208Pb(n,7)
and209Bi(n,7)
T h e termination of the s-process at 2 0 9 B i , due t o the a-decay of 2 1 0 > 2 1 1 p o a n d of 2 1 1 B i , drives t h e basic interest in t h e capture cross sections for 2 0 9 Bi a n d for t h e P b isotopes. T h e capture cross section for 2 0 4 P b , on t h e other h a n d , is essential for the analysis of the s-process branching at 2 0 4 T 1 . Overall, t h e c a p t u r e cross sections for P b isotopes are essential for t h e understanding of the nucleosynthesis of P b , in particular for the evaluation of its non-radiogenic component. From t h e experimental point of view, this set of measurements suffers from the same drawback mentioned in t h e description of the Mg isotopes: t h e extremely small a1/aei rates. It is planned t o detect are the strengths of some of t h e keV-range neutron resonances of P b and Bi. In addition to the nuclear astrophysics motivation, it is perhaps useful t o mention here t h a t the capture cross sections of P b and Bi, in a wide energy region, are requested for the development of accelerator-driven systems (ADS) for nuclear waste transmutation. Some of these nuclear devices are designed to use a Pb-Bi eutectic as neutron spallation source as well as coolant. T h e experimental determination of capture strengths for P b and Bi are clearly basic requirements for the ADS design.
3
Conclusion and perspectives
W i t h the description of the cross section measurement campaign of activities at C E R N n_TOF, we have shown how t h e characteristics of a neutron spallation source can be favorably used for neutron cross section measurements of interest in nuclear astrophysics. As a concluding remark, we mention here t h a t the construction of a 47r, high-efficiency 7-ray detector, based on B a F 2 crystals is under way at n _ T O F . This detector will allow for a considerably larger number of measurements, on even smaller samples. This represents a challenge of considerable importance with respect t o the need for more accurate measurements of neutron c a p t u r e cross sections for t h e quantitative understanding of stellar nucleosynthesis a n d advanced stellar modelling.
31 Acknowledgments This work is the result of the activity of the n_TOF Collaboration. This Collaboration is composed of approximately 130 scientists belonging to about 30 different research Institutions from Europe and the USA. The interest and enthusiasm of all these colleagues for the project and for the activities related to nuclear astrophysics is acknowledged. References 1. C. Rubbia et al., A high resolution spallation driven facility at the CERNPS to measure neutron cross sections in the interval from 1 e V and 250 MeV, CERN/LHC/98-02(EET), Geneva, May 30, 1998. 2. U. Abbondanno et al. Study of background in the measuring station at the n.TOF facility at CERN, CERN/SL-Note 046, 2001. 3. R. R. Winters, R. L. Macklin, Phys. Rev. C 25, 208 (1982).
RECENT ASTROPHYSICS RESULTS FROM ORELA AND POSSIBLE FUTURE EXPERIMENTS AT ORELA AND SNS P. E. KOEHLER Physics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831 E-mail: [email protected] I present some recent results from experiments at the Oak Ridge Electron Linear Accelerator (ORELA) and discuss their impact in nuclear astrophysics. I then describe some possible future nuclear astrophysics experiments at ORELA and at the Spallation Neutron Source (SNS) being built in Oak Ridge. The SNS and ORELA are complementary, world-class facilities and both will be needed for important future experiments in nuclear astrophysics.
1
Introduction
The ORELA facility [1] has a long and distinguished history of experiments in nuclear astrophysics. Most of the neutron capture reaction rates used in nuclear astrophysics calculations were determined in experiments at ORELA by Dick Macklin and collaborators. More recently, an improved apparatus [2] has made it possible to measure these rates much more accurately and to lower energies than before. These new data are needed to make use of new high precision isotopic anomaly data from meteorites [3] and to test the latest stellar models [4]. Also, it was recently realized [5] that («,a) experiments at ORELA could provide perhaps the best constraints on the many (y,a) rates needed for explosive nucleosynthesis calculations. In addition, ORELA would be an excellent facility for several other types of important nuclear astrophysics experiments, such as inelastic neutron scattering, neutron capture on long-lived radioactive samples, or total cross section measurements on shorter lived radioactive samples. Most of these topics are discussed in an ORELA "White Paper" that is available at http://www.phv.ornl.gov/astrophvsics/nuc/neutrons/whitepaper.pdf. The intense neutron flux at the SNS [6] should allow measurements on much smaller samples than is possible at ORELA. A recent study [7] indicates that the flux at the SNS should be over 10000 times higher that at the ORELA "benchmark" facility where most previous neutron capture measurements for nuclear astrophysics have been made, and over 40 times higher than at the new DANCE instrument (see John Ullmann's paper in these proceedings) at the Los Alamos Neutron Science Center (LANSCE) [8]. Therefore, the SNS should make possible measurements on widest range of radioactive and very small stable samples of interest to nuclear astrophysics.
32
33
2
Recent ORELA Results and the Need for New Measurements
If you are not familiar with this field, I urge you to read the excellent overview paper by Franz Kappeler at the beginning of these proceedings to acquaint yourself with the basic concepts and jargon. 2.1
The cool, new s-process models
The latest, most realistic, and most successful models [4] of the s process indicate that roughly half of the abundances of nuclides heavier than A=100 were made in low-mass Asymptotic Giant Branch (AGB) stars. A major difference between these and previous models is that most of the neutron exposure driving the nucleosynthesis occurs at much lower temperatures (kT=6-8 keV) than previously thought (kT=30 keV). This could be a problem because most of the old measurements, and even some of the new high-precision measurements, were not made to low enough energies to obtain the reaction rates at these new lower temperatures without resorting to extrapolations. At ORELA, we can routinely measure the neutron capture cross sections across the entire range of energies needed. In all [9,10] but one [11] of the cases studied so far, new ORELA measurements indicate that extrapolations from previous data to obtain reaction rates at the low temperatures needed by new stellar models are in error by two to three times the estimated uncertainties. Therefore, extrapolated rates are not sufficiently accurate for meaningful tests of new stellar models. More low-energy measurements are needed, especially for the s-only isotopes that serve as the most important calibration points for the models. 2.2
Cracks in the Classical s-process Model
The so-called classical model of the s process has been used for many years because, by making some simplifying assumptions (constant temperature, neutron density, and matter density), it is possible to find an analytical solution to the large network of time-dependent, coupled differential equations describing the reaction flow during the s process. As a result, the classical model has been very useful for ascertaining the mean conditions of the s-process environment. Although stellar models indicate that the classical model assumptions are too severe, it has been amazingly successful in reproducing the observed s-process abundances. The competition between neutron capture and beta decay at several relatively long-lived radioactive nuclides along the s-process path can yield a very direct handle with which to estimate the average neutron density, temperature and matter density in the stellar plasma during the s process. Because we know assumptions of the classical s-process model are too simplistic for real stars, if (n,y) cross-section
34
data of sufficient accuracy exist, classical analyses of different branchings should eventually yield inconsistent results. Thanks to precise new data from ORELA, cracks in classical model are beginning to show. Previous classical analyses of branchings in the ^-process path had led to a temperature of kT= 29±5 keV. In contrast, recent precise l34'l36Ba(«,y) reaction rate measurements from ORELA [8] were used in a classical analysis of a different branching to deduce a mean .y-process temperature of kT= 15±5 keV. This was the first time that clearly inconsistent temperatures were obtained from different .^-process branchings. Even more recent ORELA measurements on isotopes of Pt [12] have yielded a second example of an inconsistency, this time in the neutron density. Because stellar models are complicated, more precision measurements near other branching points (e.g., 85Kr, 95Zr, 151Sm, 152Eu, ,53Gd, 163Ho, 170'17,Tm,..) will be needed to understand the crucial ingredients in the new stellar models. Additional cracks in the classical s-process model appeared when new ORELA data [12,13] demonstrated for the first time that the classical model of the s process fails to predict the correct abundances of the ,?-only isotopes Nd and Pt. 2.3
Red Giant Stardust
Microscopic grains of silicon carbide and other refractory materials recovered from primitive meteorites represent a new class of observational data with which to constrain astrophysical models. Most of these grains appear to be actual Stardust from red giant (AGB) stars inside of which the s process had occurred. Trapped within these grains are trace amounts of several intermediate- to heavy-mass elements having isotopic patterns that are very non-solar. Qualitatively, these patterns agree with the expectations of nucleosynthesis from the s process - they are relatively enriched in .^-process isotopes and depleted in isotopes thought to come from the r and p processes. Also, the precision to which these isotope ratios can be measured is much higher than the precision of measured element-to-element abundances for the solar system. Therefore, these new meteorite data extend both the number and precision of the calibration points for .y-process models. New highprecision neutron capture data are needed to see if this beautiful, qualitative redgiant Stardust model can be made quantitative. The first precise test of the red-giant Stardust model recently was made possible when new ' Nd(n,y) cross sections were measured [13] with good precision at ORELA. Stellar ^-process model calculations made with previously accepted cross sections were in serious disagreement with the Stardust data. The new ORELA measurements, which were made with an improved apparatus and over a wider energy range, showed that the old data were in error. With the new ORELA data, the agreement between the stellar model and the Stardust data was excellent. Subsequent ORELA measurements for isotopes of barium [9,11] have revealed problems in the red giant Stardust model. Stardust data for other elements exist (e.g.,
35
Sr, Mo, and Dy), but because many of the existing (n,y) data are too imprecise or do not cover the entire energy range needed by the models, new measurements are needed to make use of these data to test and improve the red-giant Stardust model. 2.4
Improving Reaction Rates for Explosive Nucleosynthesis Models
The neutron-deficient isotopes (the so-called p isotopes) of intermediate- to heavymass elements cannot be made by neutron capture reactions starting from stable "seed" nuclides. It is thought that they were synthesized when seeds built up by a previous s process were photo-eroded in an explosive, high-temperature environment during the p process. The site of the p process is unknown, but the leading candidates appear to be the late stages in the lives of massive stars or supernova explosions. The largest nuclear physics uncertainties in these models are the rates for (y,a) reactions. Determining these rates through direct laboratory measurements is very difficult if not impossible because the cross sections are extremely small due to high Coulomb barriers. Because the level densities are high at the excitation energies and masses of interest, the rates can, in principle, be calculated to sufficient accuracy using the statistical model. However, the a+nucleus potential needed for this model is very poorly constrained, so calculated rates are very uncertain. Traditional methods for constraining potentials are problematical because they involve a largely unconstrained extrapolation. A series of low-energy (n,a) measurements, across a wide range of masses, appears to be the best means of constraining the a+nucleus potential and thus improving the calculation of these rates. The Q values for these (w,a) reactions are such that the relative energies between the a particle and the residual nucleus are in the astrophysically interesting energy range, so no extrapolation is necessary At ORELA, the first application of this idea was to measure the 143Nd and 147 Sm(n,a) [5] cross sections across the range of energies needed for astrophysics applications. Previous measurements of this type were limited to energies below a few keV (which is too small of an energy range to be useful for comparison to statistical models) due to overload problems in the detectors and associated electronics resulting from the y flash at the start of each neutron pulse. In the new experiments, this problem was overcome by employing a compensated ion chamber (CIC) [14] as the detector. The CIC reduced the y-flash background to the point where measurements are possible to much higher energies (500 keV in the cases of 143 Nd and 147Sm). Results from the ORELA measurements are shown in Fig. 1. The older calculations of Holmes et al. [15] are much closer to the data than the newer NONSMOKER [16] or MOST [17] calculations, which differ from the data by about a factor of 3 in opposite directions. The better agreement of the older model may be
36
due to a fortuitous cancellation of effects. The newer models employ a neutron potential that is known to be more reliable in this mass region. In addition, the authors of the newer models have attempted to reduce the reliance on empirical fine tuning and to take advantage of the latest physics knowledge in an effort to increase the reliability of the models away from the valley of stability. In the case of the a potential, several parameters are needed to account for the mass, energy, and nuclear structure effects. At present, the values of these parameters in the astrophysically relevant range are poorly constrained by experiment. 10J
'Nd(n,cx)
X •°jo
!
O
ORELA E x p e r i m e n t
^
MOST'3.7 NON-SMOKER/2.7 H o l m e s e l a/.'1.02
147
if
O
K
**k
+<
S m (n,<x)
ORELA Experiment MOST'2.7 NON-SMOKER/3.3 H o l m e s e f a/.M .2
"& %
*
Hi*f%i
101
E. ( k e V )
a 10"
101
102
E„ ( k e V )
Fi Figure 1. Cross sections for the Nd and 7Sm(n,a) reactions in the unresolved region. Shown are the recent ORELA measurements and the calculations of Holmes et al. [15], as well as calculations made with the newer statistical model codes NON-SMOKER [16] and MOST [17]. Note that the theoretical calculations have been normalized by factors given in the legends.
We have studied [5] the sensitivity of calculated (n,a) cross sections to the a potential and level densities employed in the model. We found that differences of about a factor of 30 could be accounted for in the variation of the potential alone. On the other hand, different level density prescriptions changed the cross section by a factor of about 1.5, far smaller than the effect of the a potential. More (w,a) data across as wide a range of masses and energies as possible are needed to constrain the several parameters thought to be necessary to define the global a potential needed for astrophysics applications. Counting rate estimates based on these initial experiments indicate that as many as 30 measurements should be possible across the mass range from S to Hf. However, a new detector that allows higher pressures and voltages, as well as more sample plates will be required for many of these measurements, and some will likely require higher flux than is available at ORELA.
37
2.5
Reaction Rates for Thermally Populated Excited States
The (n,y) reaction rates inside the thermal plasma of a star can be significantly different from the rates measured in the laboratory due to reactions involving thermally populated excited states. These stellar enhancement effects cannot be directly measured, but can be determined by measuring neutron inelastic cross sections to the same levels populated in the stellar environment. The enhancement of stellar reaction rates due to this effect can be as large as 30%, and enhancements calculated by various nuclear statistical models [15-17] can differ by substantial amounts. Such large and uncertain effects are particularly troublesome for ,?-only isotopes because they are the main calibration points for .y-process The stellar enhancement factor (SEF) for one nuclide along the s-process path ( Os) was determined [18] through neutron inelastic scattering measurements at ORELA several years ago. This measurement was particularly difficult because the excitation energy of the first excited state in 1870s is only 9.8 keV; hence, it was not possible to detect the de-excitation y rays directly and it was difficult to resolve the inelastically scattered neutrons from the larger elastic group. However, by using a clever technique that exploited the excellent time-of-flight resolution available at ORELA, it was possible to measure this cross section. There are four s-only isotopes (154Gd, 160Dy, 170Yd, and 176Hf) in addition to 187 0 calculated to have SEFs greater than 10%, and there are substantial differences between the SEFs calculated by different statistical models. In contrast, current techniques can determine laboratory reaction rates to 1-3% accuracy and isotopic abundances can often be measured with part-per-thousand accuracy. A program of («,«') measurements for these isotopes is clearly needed to determine enhancements for s-only isotopes and to improve statistical models so reliable enhancements can be calculated for other nuclides of interest. Measurements should be easier for these four s-only isotopes than for 187Os because both excitation energies and natural abundances are higher. There are about 25 other nuclides along the s-process path calculated to have SEFs larger than 10%. First excited state energies range from 8.4 keV (169Tm) to 100 keV (182W). For those with the smallest excitation energies it appears as if a technique similar to that used in the 1870s experiment will be necessary. However, it should be possible to use a flight path about a factor of four shorter (and hence have a higher counting rate or use smaller samples) than in the previous measurement [18] and still obtain resolution sufficient to separate the elastic and inelastic groups. 3
The Need for the SNS
Measurements on radioactive samples and on stable samples having very small natural abundances or small cross sections are needed for several reasons. First, measurements on radioactive samples are needed for improving models of the s
38
process. For example, neutron capture measurements for radioactive branching points along the .s-process path could greatly aid in understanding dynamics of the s process environment. Almost none of these measurements have been made. Branching points in the ^-process path for which measurements are needed include 85 Kr, IH135 Cs, 147Nd, 147'148Pm, ,51Sm, 152Eu, 153Gd, 163Dy, l63'164Ho, 169Er, 170',71Tm, 176 Lu, 185W, and ,86Re. Second, there are several lighter nuclides whose abundances are modified by («,y), (w,oc), and (n,p) reactions during the * process or during explosive nucleosynthesis. In some cases, these nuclides are of interest to y-ray astronomy (e.g., 22Na and 26A1), to meteoric anomalies (e.g., Si, CI, Ca, 50V, and Ti), or to the origin of rare isotopes of lighter nuclides (e.g., 36C1, 37'3 Ar). Although measurements exist for many of these cases, the data from different measurements are in serious disagreement, or of poor precision or questionable quality, or cover too limited an energy range for astrophysics applications. Third, very few measurements of («,y) reaction rates for nuclides involved in the p process have been made, so theoretical rates are used in p-process calculations. From measurements made near the valley of stability, it is known that calculated («,y) rates using global statistical models are accurate to within a factor of two. However, it is not known how reliable the theoretical rates are when extrapolated away from the valley of stability. Measurements on neutron deficient radioactive samples and on the very rare p-isotopes themselves are needed to improve reliability of theoretical rates that must still be used for the many cases that cannot be measured. Radioactive nuclides of interest to the p process include 53Mn, 55Fe, 57Co, 59 Ni, 9l'92Nb, 93Mo, 97Tc, 109Cd, 137La, ,39Ce, 143'l45Pm, 145'l46Sm, 148',50Gd, ,54',59Dy, 157 Tb, ,72 Hf, 195 Au, ,94 Hg,and 202 Pb. Fourth, measurements on neutron rich radioactive isotopes are needed to improve models of the r process. At the end of the r process when reactions freeze out as the temperature and neutron density decline, («,y) reactions could help smooth out the abundance distribution and improve agreement between astrophysical models and observed r-process abundances. The half lives of the involved isotopes are too short for direct measurements, but («,y) measurements as far off the valley of stability as possible would be very helpful for improving the reliability of theoretical rates. For this reason, reaction rate measurements on the following nuclides are needed: 90Sr, 123Sn, l26Sn, ,27mTe, ,82Hf, 2l0 Pb, 226Ra, and 227Ac. Minimizing the sample size needed is usually the most important consideration for measurements on very rare stable or on radioactive isotopes. The widest range of measurements of this type should be possible at the SNS because the flux at the SNS is expected to be much larger than at any other facility. High peak flux can also be important in overcoming the background from the decay of the sample, and the SNS also is expected to have the highest peak flux. Recent l91'193Ir(«,y) [19] and l71 Tm(n,y) [20, Ullmann et al., these proceedings] experiments at LANSCE have demonstrated the potential of the high fluxes available at spallation sources for
39
measurements on very small samples. For example, the Ir measurements were made with approximately 1 mg of sample and required a total of only one day of beam time each. Recently, a comparison [7] was made of various white neutron sources for measurements using radioactive samples. In Table 1, LANSCE [8], CERN-TOF [21], and the SNS [6] are compared to ORELA [1] operating under conditions (8 kW power, 8 ns electron pulse width and 40 m flight path length) typically used in previous («,y) measurements. Neutron capture measurements at the ORELA facility have also been made at flight paths as short as 10 m and the facility has run for extended periods at powers as high as 50 kW; hence, these conditions are also included in Table 1. The LANSCE facility has reliably operated at a power of 64 kW and a flight path length of 20 m has been chosen for the new DANCE instrument [20]. The CERN-TOF facility was originally envisioned with detector stations at 80 and 230 m although currently an original flight path length of 180 m is being instrumented. Under the most optimistic conditions, the power of this facility will be 45 kW. For this comparison, a flight path length of 20 m was chosen for the SNS because this would yield a time-of-flight resolution equivalent to previous nuclear astrophysics measurements at LANSCE. As can be seen in Table 1, the flux at the SNS is expected to be about 11000 times larger than at the ORELA benchmark facility. Furthermore, the peak flux at the SNS should be larger by a factor of 180000. Table 1. Ratios to the benchmark facility (ORELA at 8 kW power, 8 ns pulse width, and 40 m flight path length) and estimated samples sizes.
Parameter Flight Path Length3 (m) Power Flux at 30 keV Integral Flux 1 - 300 keV Peak Flux 1 -300 keV l/(Pulse Width) Sample Sizeb (mg)
Ratio to ORELA Benchmark LANSCE ORELA CERN-TOF 40 20 230 80 10
SNS 20
6.2 100 100
1 1 1
8 280 230
5.6 10 9.5
5.6 85 75
250 12000 10500
210
1
12000
520
12000
180000
0.083 0.25
1 25
0.032 0.10
6.7 2.5
2.3 0.31
0.011 0.0022
a The flight path length is given to define the facility parameters. All other entries except sample sizes are ratios to the benchmark facility. b The samples sizes are scaled to a cross section of 1 b and an atomic number of 150, and assume a detector efficiency of 100%.
40
Estimates of the sample sizes needed for («,y) measurements are also given in Table 1. The sample sizes have been scaled from previous measurements at the benchmark facility assuming a Maxwellian-averaged cross section at kT=30 keV of 1 b, an atomic number of 150, and a 100% efficient detector. These general numbers can be used to estimate the sample size needed for a particular case by dividing by the calculated 30-keV cross section in b and scaling to the atomic mass. These estimates indicate that it should be possible to measure many of the («,y) reaction rates for radioisotopes of interest to astrophysics at the SNS. In addition, the sample sizes are small enough that the necessary isotopically enriched samples should be affordable for many of the p isotopes. Finally, it should also be possible to measure most of the (n,oc) cross section of interest to explosive nucleosynthesis. Although the SNS appears to have much potential for future nuclear astrophysics experiments, there are likely to be a number of challenges to overcome. Producing the radioisotopes and fabricating the radioactive samples will present a significant challenge. In addition, it likely will be difficult to build a detector capable of recovering from the "y-fiash" at the beginning of each neutron pulse so that measurements can be made to the relatively high energies (few hundred keV) needed. Also, overcoming the backgrounds from the radioactive sample, from neutrons scattered from the sample and its backing, and from other flight paths is expected to be a challenge. Simulations [22] indicate that a detector based on the scintillator BaF2 together with judicious layers of neutron absorbing materials appears to be the best choice. SNS capabilities are complementary to ORELA and both will be needed to satisfy all the data needs for nuclear astrophysics. For example, in addition to the needed measurements outlined above, ORELA is ideally suited for measuring total cross sections using small amounts of radioisotopes [23]. Even with the large flux available at the SNS, («,y) cross sections for some radioisotopes will remain out of reach for which total cross section measurements should be possible at ORELA. These data could be very useful in constraining statistical model calculations of the («,y) rates. Acknowledgements ORNL is managed by UT-Batelle, LLC for the U.S. Department of Energy under contract number DE-AC05-00OR22725. References 1. R. W. Peele et ai, Technical Report No ORNL/TM-8225, Oak Ridge National Laboratory; K. H. Bockhoff etal, Nucl. Sci. Eng. 106, 192 (1990).
41
2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13. 14. 15. 16.
17. 18. 19.
20.
21. 22. 23.
P. E. Koehler et al., Phys. Rev. C 62, 055803 (2000). See for example, G. K. Nicolussi et al., Phys. Rev. Lett. 81, 3583 (1998). See for example, C. Arlandini et al., Astrophys. J. 525, 886 (1999). Yu. M. Gledenov et al., Phys. Rev. C 62 042801(R) (2000). D. Olsen et al, Technical Report, Oak Ridge National Laboratory Report SNS 100000000PL0001R00, (1999). P. E. Koehler, Nucl. Instr. and Meth. A460, 352 (2001). P. W. Lisowski, C. D. Bowman, G. J. Russell, and S. A. Wender, Nucl. Sci. Eng. 106,208(1990). P. E. Koehler et al., Phys. Rev. C 54, 1463 (1996). P. E. Koehler et al., Phys. Rev. C 64, 065802 (2001). P. E. Koehler et al., Phys Rev. C 57 R1558 (1998). P. E. Koehler et al., to be published in The Proceedings of the International Conference on Nuclear Data for Science and Technology, Oct. 7-12, 2001, Tsukuba, Japan. K. H. Guber, R. R. Spencer, P. E. Koehler, and R. R. Winters, Phys. Rev. Lett. 78,2704(1997). P. E. Koehler, J. A. Harvey, and N. W. Hill, Nucl. Instr. and Meth. A361, 270 (1995). J. A. Holmes, S. E. Woosley, W. A. Fowler, and B. A. Zimmerman, At. Data Nucl. Data Tables 18, 305 (1976). T. Rauscher and F.-K. Thielemann, in Stellar Evolution, Stellar Explosions, and Galactic Chemical Evolution, Edited by A. Mezzacappa (Institute of Physics, Bristol, 1998), p. 519. S. Goriely, in Nuclei in the Cosmos, edited by N. Prantzos and S. Harissopulos (Editions Frontieres, Gif-sur-Yvette, 1998) p. 484. R. L. Macklin, R. R. Winters, N.W. Hill, and J. A. Harvey, Astrophys. J. 274, 408(1983). P. E. Koehler and F. Kappeler, in International Conference on Nuclear Data for Science and Technology, edited by J. K. Dickens (American Nuclear Society, La Grange Park, 1994) p. 179. J. L. Ullmann et al., in Proceedings of the Fifteenth International Conference on the Application of Accelerators in Research and Industry, edited by J. L. Guggan and I. L. Morgan (American Institute of Physics, New York, 1999) p. 251. C. Rubbia et al., Technical report, European Laboratory for Particle Physics Report CERN/LHC/90-02 (EET), 1998; CERN/SPSC 99-8 SPSC/P 310, 1999. M. Heil, etal, Nucl. Instr. Meth. A459, 229-246 (2001). R. W. Benjamin et al., Nucl. Sci. Eng. 85, 261 (1983).
SENSITIVITY OF ISOTOPE YIELDS TO REACTION RATES IN THE ALPHA RICH FREEZEOUT
G. C. JORDAN IV AND B. S. MEYER Department of Physics & Astronomy, Clemson University, Clemson, SC E-mail: [email protected]
29634-0978
The alpha-rich freezeout from nuclear statistical equilibrium occurs during type II (corecollapse) supernovae when the shock wave passes through the Si shell of the star. The nuclei are heated to high temperature and broken down into nucleons and alpha particles. These subsequently reassemble as the material expands and cools. The alpha-rich freezeout is responsible for a number of important supernova observables. In this paper we introduce a web-based tool that displays the results of a reaction-rate sensitivity study of alpha-rich freezeout yields. This tool permits the user to identify nuclear reactions that govern the synthesis of important observables from the alpha-rich freezeout. The tool is intended to aid in the identification of nuclear reaction rates important for measurement.
1
INTRODUCTION
Of all the astrophysical settings in which nucleosynthesis takes place, the supernova is one of the most significant. Though there are several different "types" of supernovae, we wish here to consider core-collapse supernovae. During the core collapse event of a dying massive star, a shock wave created by the rebound of the collapsing material heats the surrounding material. The matter then cools and expands very quickly. In the innermost regions of the ejecta, the nucleosynthesis that takes place under these extreme conditions has been dubbed the alpha-rich freezeout due to the high abundance of alpha particles remaining at the end of the process. In this paper we introduce a web site that displays data from a sensitivity calculation performed on an alpha rich freezeout model. The purpose of the sensitivity study is to identify nuclear reactions that are important in the production of nuclei that are identified as astrophysical observables. With a particular isotope in mind, one may then use the web site to view the effect of different reaction rates on the yield of that isotope. If an isotope is sensitive to a particular reaction cross section, then improving our knowledge of that cross section would improve the model and any insights gained from the model. The web site is thus ultimately intended to spark the experimental determination of reaction rates that govern astrophysical observables. It is worth reiterating here the four requirements that should be satisfied such that the measurement of a reaction rate is warranted [1]: 1. An appropriate astrophysical model of a nucleosynthesis process must exist.
42
43
2. 3. 4.
An observable from that process, usually an abundance result, is either known or measurable. The dependency of the value of the observable on the value of the nuclear cross section is demonstrable. An experimental strategy for measuring the reaction rate, or at least using experimental data to better calculate the reaction rate, should be available.
As mentioned above, meeting these requirements will ensure that measurement of the reaction cross section will have astrophysical significance. This paper begins with a description of the alpha-rich freezeout calculations and the attendant sensitivity studies. Next the interactive web site is introduced along with detailed instructions on its use. Finally two astrophysical observables (55Ni and 57Ni) are discussed in order to illustrate use of the web site and to identify their governing reaction rates. 2
THE CALCULATIONS
In order to explore the sensitivity of alpha-rich freezeout yields to variations in reaction rates, we employed the Clemson nucleosynthesis code [2], which we have updated to employ the NACRE [3] and NON-SMOKER [4] rate compilations. The network used was that of The et al. [1], which includes 376 species from neutrons and protons up to Z=35 (Bromine) and 2,141 reactions among these species. We began the survey by running a set of reference calculations. The calculations used reaction rate values from the rate compilations. Each calculation began with a temperature of T9 = T/109 K = 5.5 and a density p = 107 g/cm3. The matter was taken to expand on a density e-folding timescale of 0.2 seconds and to obey the relation p <* T 3 , appropriate for a radiation-dominated expansion. We considered three possible values of the neutron excess T] (defined as the excess number of neutrons per nucleon), namely, 0, 0.002, and 0.006. Because we considered the appropriate model for alpha-rich freezeout to be passage of a shock wave through 28Si-dominated matter, the initial composition was 28Si and enough neutrons to give the appropriate value of T). Once the reference calculations were completed, we explored the governing effect of a particular reaction by multiplying the reference reaction rate value for that reaction by a factor of ten, running a calculation with the same initial temperature and density, expansion timescales, and r\ as in the appropriate reference calculation, and recording the results. We repeated this procedure for all 2,141 reactions and for all initial r|'s. We then repeated the whole sequence, but this time for a multiplicative factor 0.1 on all the reactions. The total number of alpha-rich freezeout calculations performed, then, was 2,141 (number of reactions) x 3(number of different neutron richnesses) x 2 (factor of 10 or factor of 0.1 on reaction rates) + 3(reference calculations) = 12,849.
44
The governing effect of a particular reaction on the yield of a particular species can be determined from the ratio of yield of the species in the calculation with the modified rate to the yield in the appropriate reference calculation. Clearly with 12,849 different calculations and 376 species, there are many combinations of results to consider. Rather than present many tables, we instead load the data on a web-based data display (figure 1) to allow the interested reader to explore the results in detail on his or her own. We describe this web tool in the next section. F i.1 LI nmimiAmmiTiiiFniEic j
. •> & •
*K-i.
».-
ut
C * -1 .
S •
RAIE FACTOR
NETJIROH EXCESS
fro i f ooo? Iff 0 006 UgtoapTDataFiEe
12
13
ASCENDING LIST Fraction of Standard
|l
,t»57 + nl <—> hi 1 co57
t- 1 57680e+OS sees
[l 8043642
0 87572479
jni56 + nl <-- >hi +co56
[o 75204492
|0 99072928
'm57+nl<-- > i»58
[o 9B528887
JO 99074060
m57 + nl <-- >hc4+fe54
|o 98643829
[» 99087404
cu60 + h l < - •>he4 + ru57 ra57+hE4-:—>ai61
fa
99067951
0 99153874
fo 99605758
099635454 |0 99707891
,11156 + 111 <•
zn61+nl < - > h l + c u 6 1
|o 98797416
[0 99736149
f
co57 + h l «
•>ni58
jo 77309935
[0 99876414
ini58 + nl <
>ni59
jo 96352522
jo 99894000
lco56+hl<
->t»57
fa 99891188 h Qf.Mnt'ri
0 99909458
->5Q4
W<x»~
0091)0^7 !$lrtemtt
Fig. 1 General layout of the interactive web based application used to display the alpha rich freezeout sensitivity data. Each numbered item corresponds to a subheading in the next section. The URL of the web site is http://photon.phys.clemson.edu/gjordan/Ion/Ti44 .
3
WEB-BASED DATA DISPLAY
As mentioned above, the results of the alpha rich freezeout survey can be viewed on the World Wide Web. In this section we introduce the interactive data display and give instructions for its use. The URL of the interactive data display is http://photon.phys.clemson.edu/gjordan/Ion/Ti44/index.html.
45
To operate the web site, first select the data set you wish to view (specified by the Rate Factor and Neutron Excess options). Next, enter the isotope of interest in the text fields and click the appropriate button to generate the desired data lists. The website also includes a Beta Decay Option which takes the abundances from the alpha rich freezeout calculation and allows them to beta decay for a specified amount of time. This is useful since the astrophysical sites of the alpha-rich freezeout (type II supernovae) are often observed many years after the explosive nucleosynthesis has taken place. The Beta Decay Option allows the abundances to beta decay and thus be compared to experimental data at the time of observation. In the subsequent subsections of this section, we present an explanation for the several user inputs and possible outputs from the web-based data display.
HATE FACTOR
KEUTRON EXCESS r0 r 0.002 0.006
Uporj J Data F'le Fig. 2 The data file selection area of the web site.
3.1
Rate Factor
The Rate Factor option selects the data set that either has the individually varied reaction rate lowered by a factor of 10 or raised by a factor of 10 (figure 2). In figure 2, the choice is for an increase by a factor of 10. 3.2
Neutron Excess
The Neutron Excess option selects the data sets with the initial Neutron Excess as 0, 0.002, or 0.006 (figure 2). In figure 2, the choice is for a neutron excess of 0.006. 3.3
Upload Data File
Click this button to upload the data file once the rate factor and neutron excess have been selected (figure 2). 3.4
Atomic Number: Z
Enter the Atomic Number of the isotope of interest in the text field (Figure 3).
46
3.5
Mass Number: A
Enter the Mass Number of the isotope of interest in the text field (figure 3).
Beta Decay P Decay Time (years): | : Ascending list "•
Fig. 3 The data display controls. Use these buttons to select the desired isotope, the beta decay option and the format in which the data is to be displayed. A list of the isotopes and nuclear reactions used in the network can also be displayed.
3.6
Beta Decay Option
Click this check box to select the Beta Decay Option. The results of the initial calculation from the alpha rich freezeout simulations are allowed to beta decay for the specified number of years. The results of the beta decay are displayed when the Ascending List or Descending List buttons are clicked (figure 3). 3.7
Decay Time
Enter the amount of time in years to Beta Decay in the text field (figure 3). 3.8
Ascending List
Click this button to generate a data list of the isotope yields. The list is sorted such that the varied reaction rate resulting in the smallest Fraction of Standard is displayed at the top of the list, followed by the rest of the varied reaction rates whose Fraction of Standard are arranged in ascending order. If the Beta Decay Option is selected, the list is sorted based on the isotope yields after they have been allowed to beta decay (figure 3).
47
3.9
Descending List
This button also generates a data list of the isotope yields. The list is sorted such that the varied reaction rate resulting in the largest Fraction of Standard is displayed at the top of the list, followed by the rest of the varied reaction rates whose Fraction of Standard are arranged in descending order. If the Beta Decay Option is selected, the list is sorted based on the isotope yields after they have been allowed to beta decay (figure 3). 3.10 Reaction List Click this button to display a list of the nuclear reactions that are in the network (figure 3). 3.11 Isotope List Click this button to display a list of the species in the nuclear reaction network (figure 3).
12
13 Rank
ASCENDING LIST Reaction
ni57 + nl <—>hl + co57
Fraction of Standard
14
A5
Fraction^of Standard at t = 1.57680e+08seci 0.87572479
m 5 6 + n l < — > h l + co56
1.8043642 075204492
m57 + nl"<—>ni58
jo 9852888?"
0.99072928 099074060"
ni57 + nl <—> he4 + fe54~
JO 98643829
0.99087404
cu60 +"i»T<--^e4"+ri57
" ja9906795l"
Fig. 4 The data display area of the web site.
3.12
Rank
This column in the output frame shows the order in which a specific reaction rate appears on the data list (Figure 4). 3.13 Reaction This column shows the relevant reaction rate that is varied (either increased of decreased by a factor of 10) (figure 4).
48
3.14 Fraction of Standard This column indicates the effect that the varied rate has on the isotopic yields. The numerical value is the yield of the selected isotope produced with the indicated reaction rate (the reaction rate that appears adjacent to the Fraction of Standard value on the list) varied, divided by the yield of the selected isotope with all of the reaction rates set to their standard values. The order in which the data appears on the list depends on these values when the Beta Decay Option is not selected (figure 4). 3.15 Fraction of Standard at t—? This column appears if the Beta Decay Option has been selected. At the top of the column the time in years that was entered in the years text field is converted to seconds. The numerical value is computed in the same manner as above. The value is the isotopic yield produced with the varied reaction rate and allowed to beta decay for the set amount of time divided by the isotopic yield produced with the standard reaction rate and also allowed to beta decay for the set amount of time. The order of the data list is based on this column when it is present. 4
EXAMPLES
In this section, we present some results of our calculations drawn from the web site described above. For brevity, we will focus only on r|=0.006 calculations with the reaction rates increased by a factor of 10. The purpose is to illustrate how one may identify relevant observables and find their governing reactions. These are just examples—the interested researcher may find many other important cases by exploring the web site. At least three types of isotopic observables are relevant for stellar nucleosynthesis. First, there is the bulk yield, which is important for understanding Galactic chemical evolution and solar system abundances. Second, there is the yield of radioactive species, such as 26A1 or ^Ti, which may be observed from space telescopes and provide constraints on their production sites. Third, there are the isotopic effects for trace elements in presolar grains [5]. For our examples, we will examine two isotopes of nickel. In alpha-rich freezeouts, these nickel isotopes are either directly or indirectly linked to all three types of observables; hence, our examples will focus on them. 4.1
56
Ni
Roughly one-half the 56Fe in the solar system was produced in alpha-rich freezeouts in core-collapse supernovae [6], which makes the bulk yield of 56Fe an important observable. 56Fe is produced primarily as radioactive parent 56Ni (ti/2=6.075 days), which decays through 56Co (ti/2=77.2 days) to 56Fe. The overlying supernova largely
49
obscures gamma rays from the decay of 56Ni, but the 2.598 MeV gamma rays (for example) from 56Co can escape and be detected [7]. These gamma rays, and the supernova light curve, which is powered in its early phases by the gamma rays that do not escape, are significant observables associated with 56Ni. From the web site, one finds that the yield of 55Ni immediately after the r|=0.006 alpha-rich freezeout is quite insensitive to a factor of ten increase in any reaction rate. The largest effect is a 2.4% increase in the 56Ni yield for a factor of ten increase in the triple-alpha reaction rate. This 2.4% increase propagates to a 2.4% increase in the 56Co yield after the 56Ni has decayed and a 2.4% increase in the final yield of 56Fe. These are small effects, and we conclude the calculated yields of alpha-rich freezeout observables associated with 56Ni are quite robust against uncertainties in any nuclear reaction rates. | l l l l l i i r i |
| I I I I I I I I l-
7] = 0.OD6 Ni(n.p)57Co
1 x
•••
10 x "Ni(n,p) s 7 Co
-
0.10
57
—
0.1 x
s7
Ni(n,p) s7 Co
- QSE
\
x
0.01
i
6
5
i
t
i
i
i
i
i
4
i
i
i
i
i
i
i
i
\
\
\
\
\
\
i
3
2
1
0
Fig. 5 The mass fraction of 57Ni plotted for several values of the 57Ni(n,p)57Co reaction rate. Also plotted is the QSE mass fraction of 57Ni. Notice the reaction rate governs when 57Ni breaks out of QSE.
4.2
"Ni
A second important isotope is 57Fe. This isotope is similar to 56Fe in that it owes a significant portion of its synthesis to alpha-rich freezeouts and it is primarily produced in alpha-rich freezeouts as parent 57Ni. Gamma-rays from the decay of Co are observable with space detectors and they also power the supernova light
50
curve at somewhat later times than Co [8]. The yield of 7Ni is thus another significant observable from alpha-rich freezeouts. The web site shows that the yield of 57Ni immediately after the alpha-rich freezeout is, like that of 56Ni, largely insensitive to the value of any particular reaction rate because both are largely equilibrium products. The reaction with the most significant effect for r|=0.006 is 57Ni(n,p)57Co, which produces a 12.5% decrease in the 57Ni yield for a factor of ten increase in the rate. The primary reaction governing the 57Ni yield for r|=0.006, then, is 57Ni(n,p)57Co, and the uncertainty in this rate gives rise to a several percent uncertainty in this observable. The reason 57Ni(n,p)57Co has an effect on the 57Ni yield is that it governs when 57 Ni falls out of quasi-equilibrium (QSE [9]). As figure 5 shows, the QSE abundance of 57Ni falls after T9 = 4.3. As the matter cools further, the QSE abundances shift to higher mass nuclei. The faster the 57Ni(n,p)57Co rate, the longer the 57Ni stays in QSE and the lower its final abundance at freezeout. Conversely, decreasing the rate means 57Ni falls out of QSE earlier and retains more of its originally high abundance. 5
CONCLUSION
The above web site is intended to be a starting point for identifying important reaction rates in the production of astrophysical observables. The interested experimentalist may explore the production factors of isotopes produced in the alpha rich freezeout under several different conditions. Once a reaction rate is identified as significant, a more detailed analysis of the effect of the reaction rate can be performed. We hope that this web site will be an aid to experimentalists in identifying important nuclear reaction rates to measure. The examples listed above are only two of many observables. Already we see that uncertainties in the 57Ni(n,p)57Co reaction rate could have an effect on the predicted yield of 57Ni from the alpha rich freezeout model. With the advent of new x-ray telescopes, new nuclear x-ray lines may become detectable [9]. New isotopic data from meteorites is also being collected. These sources of data may reveal new observables, and, hence, new tests of the alpha rich freezeout models. Having accurate data on the reaction rates that govern the production of these observables will improve the models, and ultimately improve our understanding of supernovae. Acknowledgements This work was supported by NASA grant NAG5-10454 and NSF grant AST 9819877.
51 References 1. L. The, D. D. Clayton, L. Jin, and B.S. Meyer, Astrophys. J. 504, 500 (1998) 2. B. S. Meyer, Astrophys. J. Lett. 499, L55 (1995) 3. C. Angulo, M. Arnould, M. Rayet, P. Descouvemont, D. Baye, C. LeclercqWillain, A. Coc, S. Barhoumi, P. Aguer, C. Rolfs, et al., Nuclear Physics A, 656, 3 (1999) 4. T. Rauscher and F. Thielemann, Atomic Data and Nuclear Data Tables 75, 1 (2000) 5. A. M. Davis, R. Gallino, M. Lugaro, C. E. Tripa, M. R. Savina, M. J. Pellin, and R. S. Lewis, in 33rd Lunar and Planetary Sci. Conf., March 11-15, 2002, Houston, TX, abstract #2018 (2002) 6. F. X. Timmes, S. E. Woosley, and T. A. Weaver, Astrophys. J. Suppi, 98, 617 (1995) 7. M. D. Leising and G. H. Share, Astrophys. J. 357, 638 (1990) 8. D. D. Clayton, M. D. Leising, L.-S. The, W. N. Johnson, and J. D. Kurfess, Astrophys. J. Lett., 399, 141 (1992) 9. B. S. Meyer, T. D. Krishnan, and D. D. Clayton, Astrophys. J., 498, 808 (1998)
NEUTRON REACTIONS OF LIGHT NUCLEI FROM ASTROPHYSICS & NUCLEAR PHYSICS INTEREST
Y. NAGAI, T. SHIMA, A. TOMYO, H. MAKI1, K. MISHIMA AND M. SEGAWA Research Center for Nuclear Physics, Osaka Univ. Iharaki, Osaka 567-0047, E-mail :naeai&>,rcnv. osaka-u. ac. jp
Japan
M. IGASHIRA, AND T. OHSAK1, Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology, Ookayama ,Meguro, Tokyo 152-8550, Japan The measurement of the neutron capture reaction of a light nucleus at astrophysical relevant energy is important for the studies of nuclear astrophysics and nuclear physics. In the measurement one can expect to detect a discrete y-ray emitted promptly from a neutron capturing state feeding to a low-lying state of a final nucleus, since the nuclear level density of a light nucleus is low. The prompt discrete y-ray characterizes the final nucleus and therefore its detection is crucial to unambiguously determine the small neutron capture cross section of a light nucleus. In addition the discrete y-ray carries unique information for investigating the reaction mechanism and nuclear structure of a final nucleus. Hence we have developed a prompt discrete y-ray detection method using pulsed neutron beams to measure the neutron-capture cross section of a nucleus. In this paper we discuss experimental results of keV neutron capture cross section measurement of a light nucleus using our method.
1
INTRODUCTION
Fast keV neutrons have been used to measure the neutron induced reaction cross section of a nucleus and thereby to construct models of stellar evolution and nucleosynthesis in stars f 1]. Among the neutron reaction studies the capture reaction of a light nucleus plays crucial roles in the studies of primordial nucleosynthesis [2] and stellar nucleosynthesis [3], and of nuclear physics [4,5]. A decade ago inhomogeneous big bang models have been proposed as alternative models of the standard big bang models [6]. According to the inhomogeneous big bang models, an intermediate mass nucleus up to A=30 could have been produced in neutron rich regions via following neutron capture reactions such as p(n,Y)D(n,y)3H(d,n)4He(t,Y)7Li(n,Y)8Li(a,n)1 ,B(n,T)12B(Pl,2C(n,Y),3C(n.y) C etc. [7]. The prediction differs from that of the standard models, in which the elemental production up to Li has been predicted [2]. If one could observationally distinguish these two models, it would give us a deeper insight of the physics in the early universe. In order to estimate the production yield of the intermediate mass nuclei during the primordial nucleosynthesis and thereby to discriminate various inhomogeneous models it is certainly important to measure the neutron capture cross section of these light nuclei at astrophysical relevant energy. Here it should be added that since the primordial abundance of D and 4He has not yet been
52
53 determined [8], it is necessary to accurately measure the neutron capture cross section of a nucleus relevant to the production and/or destruction of the D and 4 He. Concerning problems related to stellar nucleosynthesis in evolved stars such as the sun we have reasonable models. However, we do not have proper models for heavy elements nucleosynthesis in old metal-deficient stars. Hence current efforts are being made to observe the heavy elements abundance in these old stars [9] and also in primitive meteorites (pre-solar grains) [10] to construct the models of stellar evolution and nucleosynthesis. It should be stressed that the observed abundance is considered to carry important information on the nucleosynthetic processes in the earlier Galactic stellar generations. Regarding the s-process nucleosynthesis in these old metal-deficient stars and in pre-solar grains, abundant light nuclei such as C, N, O and Ne et al. would act as a strong neutron poison [11], if the neutron capture cross section of these nuclei would be large. The production yield of heavy elements in the old stars, therefore, depends strongly on the cross section. Regarding the rprocess nucleosynthesis the astrophysical site remains as an open question to be solved. Woosley et al. have proposed the possible site in nascent neutron star winds [12]. In this scenario light nuclei would also affect the production yield of heavy elements due to a possible neutron poison. As discussed above the study of a neutron-capture reaction of a light nucleus is quite important in the field of nuclear astrophysics. Its study is also quite interesting from nuclear physics viewpoint as described below. A light nucleus is characterized to have a low level density. Hence we can expect to observe a discrete y-ray emitted promptly from a neutron capturing state feeding to a low-lying state of a final nucleus, if we can develop a high sensitive y-ray detector. The discrete y-ray carries a unique electromagnetic property such as E l , Ml and E2 etc. Therefore its observation gives us unique information of initial and final states of a final nucleus, which connects the y-ray. Through the information one can learn the neutron capture reaction mechanism and nuclear structure of the final nucleus. It could happen that such nuclear physics information is quite unique and can be hardly obtained by another experimental approaches. Here it should be mentioned about an interesting point of the study of a few body nuclei. Through the study we can pickup a small finite effect due to non-nucleonic degrees of freedom in a nucleus. A wellknown example is the thermal neutron capture by a proton [4]. The experimental value of 334 mb is about 10% larger than the calculated one based on an impulse approximation, and this value of 10% has been considered to originate from the meson exchange currents [4]. Such effect due to the non-nucleonic degrees of freedom has been also observed in the thermal neutron capture by a deuteron [13]. Currently, however, calculated values including the meson exchange currents and delta current do not agree with the measured value, and thus further extensive studies of a few body nuclei are important to reveal the role of the non-nucleonic degrees of freedom in a nucleus. The measurement of the neutron capture cross section at stellar energy plays an important role, since the low energy reaction
54
makes the reaction mechanism simple and thus one can obtain a reliable theoretical value. Here it should be stressed that the detailed study of the neutron capture reaction of a stable light nucleus is quite important to understand the reaction mechanism and nuclear structure. The understanding is essential to estimate the neutron capture cross section of an unstable nucleus at stellar energy for calculating the production yield for primordial intermediate mass nuclei and r-process nuclei. 2
EXPERIMENTAL METHOD
The measurement of keV neutrons capture reaction cross section of a nucleus has been made by two different methods of activation and prompt y-ray detection [14]. We have been using the latter method, in which an anti-Compton Nal(Tl) spectrometer has been developed for detecting a prompt discrete y-ray emitted from a neutron capturing state feeding to a low-lying state of a final nucleus [15]. Here it should be mentioned that a discrete y-ray detection is crucial to identify a true signal from the neutron capture reaction of a target nucleus free from background and determine a small capture cross section of a light nucleus unambiguously. This is because discrete y-rays characterizes a final nucleus. Our experimental method is described briefly. Pulsed keV neutrons are produced by the 7Li(p,n)7Be reaction using pulsed protons provided from the 3MV Pelletron Accelerator of the Research Laboratory for Nuclear Reactors at the Tokyo Institute of Technology. Discrete prompt y-rays from a capturing sample are detected by a high sensitive antiCompton Nal(Tl) y-ray spectrometer [16]. Here, the compound 6LiH inside the envelope of the spectrometer effectively attenuates intense neutrons scattered by a sample without emitting any y-ray. The spectrometer is placed at 125.3° with respect to the proton beam direction. Hence the y-ray intensity measured at this angle gives an angle-integrated y-ray intensity for the dipole transition, since the second Legendre polynomial is zero at this angle. A gold sample has been used for normalization of the cross section, since the capture cross section of gold is well known. The (n,p) and/or (n,oc) reaction of a light nucleus also plays important roles in stellar nucleosynthesis. Both the reaction cross section and energy of emitted charged particles induced by low energy neutrons are low. Hence we have developed a new detector, gas scintillation drift chamber (GSDC) [17], which can detect low energy charged particles with a large solid angle of 4n and high efficiency of 100 %. The detector consists of two detection parts of primary scintillation photons and ionized electrons. The photons are detected by four photomultipliers, which are attached to the chamber, and are used as start signals to determine the neutron energy by a TOF method. The electrons are detected by multi-wire proportional counters in the GSDC. The high sensitivity of the GSDC for detecting low-energy charged particles emitted from the keV neutron-induced
55
reaction of a nucleus has been shown for the measurement of the reaction cross section [18]. 3
14
N(n,p)' C
RESULTS AND DISCUSSIONS
The measured results of the neutron capture reaction cross section of a light nucleus are briefly discussed below. The cross section of the p(n,y)D reaction has been successfully measured at several neutron energies between 10 and 200 keV and 550 keV with an uncertainty of 5-10 % [5]. The obtained cross sections are in good agreement with the value by Hale et al. [19] and theoretical values by Sato et al., who take into account the meson exchange currents [20]. In this measurement a newly developed Monte Carlo code, TIME-MULTI, played an important role to correct for the effect of the multiple scattering of an incident neutron with a hydrogen in a polyethylene sample [21]. Here it should be mentioned that the p(n,y)D reaction cross section estimated by Hale et al. has been used for calculating the yield of primordial light elements. Through the detailed studies of the neutron capture reaction of 12C and l 6 0 we have observed a non-resonant direct p-wave capture reaction process, which enhances significantly the cross section even at astrophysical relevant energy of a few 10 keV [22]. The measured capture cross section of 16 0 at the average energy of 30 keV is 34 ubam, which is about 170 times larger than the estimated value using a 1/v law. The astrophysical impact of this large value has been considered in the nucleosynthesis of heavy elements in massive low metallicity stars [23], which claims that 16 0 strongly reduces the production yield of elements heavier than iron as the metallicity decreases. The large cross section has been discussed from the nuclear physics interest. Its value is explained as a new process, called a nonresonant direct p-wave capture reaction, which populates directly a 1 12 (si/2) state in 17 0 [22]. Here it should be noted that the l+/2 (si/2) state has a large spectroscopic factor of about 1 with relatively low neutron binding energy. Hence incident neutrons can be captured directly by 16 0 by transferring an orbital angular momentum of 1 to the final nucleus 17 0 and one sees a non-resonant direct p-wave capture process [24]. Here it is worthwhile to mention about a new experimental method to measure a neutron capture reaction of an unstable nucleus. The method utilizes the Coulomb dissociation of radioactive ion beam, and it has been used to simulate the nucleosynthesis in stars [25]. The dissociation reaction is inverse reactions of a neutron capture reaction of a target nucleus. It has been claimed that by a proper selection of the kinematical condition of the reaction a non-resonant direct p-wave capture process has been observed in the dissociation reaction. Although the method is quite unique, both the dissociation cross section and intensity of the RI beam are low. Hence it has been applied for an unstable nucleus, which has an extremely low neutron binding energy of below 1 MeV and a ground state spin of 1/2 (S1/2). In
56
order to clarify the second point we compare the experimental result of the Coulomb dissociation of 19C to that of the (n,y) reaction of 16 0 in figure 1. In the 16 0(n,y)170 reaction a non-resonant direct p-wave capture process populates strongly the excited state l/2+ (s1/2) in addition to the week population of the ground state 5/2+ (d5/2) by emitting y-rays of above 3 MeV. Namely, the direct neutron capture reaction is more general compared to a Coulomb dissociation method. Therefore it is very important to measure the neutron capture cross section of an unstable nucleus with a proper long half-life.
0.530 MeV
18.
1/2+
19
17
o
(b)
(a) 19
Figure 1. Coulomb dissociation of C (a) is compared to the direct (n,y) reaction of l7 0 (b). The latter reaction strongly populates the excited l/2+ (si/2) state in , 7 0 . 4
SUMMARY
The measurement of the neutron induced reaction cross section of a light nucleus at stellar energy has been successfully made by developing high sensitive y-ray and charged particle detectors system. It has been experimentally shown that the detection of a discrete y-ray, which connects between initial and final states of a residual nucleus, is crucial to determine a small capture cross section accurately and thus to obtain detailed information of the reaction mechanism and nuclear structure. Since many important and interesting problems of the neutron induced reaction of light nuclei remain to be attacked both experimentally and theoretically, it would be extremely nice to extend such studies using intense spallation neutrons. References 1. Clayton D. D., Principles of Stellar Evolution and Nucleosynthesis (The University of Chicago Express, Chicago and London, 1983)
57
2. Wagoner R. V., Fowler W. A. and Hoyle F., Astrophys. J. 148 (1967) 3 3. Burbidge E., Burbidge G., Fowler W. and Hoyle F., Rev. Mod. Phys. 29 (1957) 547, Gallino R. et al., Astrophys. J. 334 (1988) L45 4. Riska D. O. and Brown G. E., Phys. Lett. 38B (1972) 193 5. Nagai Y. et al., Phys. Rev. C56 (1997) 3173, Kikuchi T. et al., Phys. Rev. C 57 (1998) 2724 6. Applegate J. H., Hogan C. J., and Scherrer R. J., Phys. Rev. D 35 (1987) 1151 7. Malaney R. A., and Fowler F., Astrophys. J., 333 (1988) 14 8. Songaila A. et al. Nature 368 (1994) 549, Olive K. A. et al, Astrophys. J, 482 (1997) 788, Izotov Y. I. Et al, Astrophys. J , 500 (1988) 188 9. Sneden C. et al., Astrophys. J. 467 (1996) 819 10. Anders E. and Zinner E , Meteoritics 28 (1993) 490 11. Prantzos N , Hashimoto M , and Nomoto K, Astronomy & Astrophysics, 234 (1990)211 12. Woosley S.E. and Hoffman R.D, Astrophys. J. 395 (1992) 202 13. Carlson J. and Schiavilla R. et al. Rev. Mod. Phys. 70 (1998) 743 and references therein. 14. Kaeppeler F , Progress in Particle and Nuclear Physics 43 (1999) 419 15. Igashira et al., Nucl. Instr. Meth. A 245 (1986) 432, Nagai Y. et al., Astrophys. J. 372 (1991) 683 16. Ohsaki et al., Nucl. Instr. Meth. A 425 (1999) 302 17. Shima T. et al., Nucl. Instr. Meth. A 356 (1995) 347 18. KiiT. et al., Phys. Rev. C 59 (1999) 3397 19. Hale G. M. et al., ENDF/B-VI Evaluation, Material 125, Revision 1 20. Sato T, Niwa M. and Ohtsubo H, in Proceedings of the Int. Symp. on weak and Electromagnetic Interactions in Nuclei, edited by Ejiri H, Kishimoto T, and Sato T. (World Scientific, Singapore, 1995) 488 21. Senoo K. et al., Nucl. Instr. Meth. A 339 (1994) 556 22. Nagai Y. et al., Astrophys. J. 372 (1991) 683, Ohsaki T. et al, Astrophys. J. 422 (1994) 912, Igashira M. et al., Astrophys. J. 441 (1995) L89 23. Rayet M. and Hashimoto M, Astronomy and Astrophysics. 354 (2000) 740 24. Ho Y. K, Kitazawa H, and Igashira. M, Phys. Rev. C 44 (1991) 148, Mengoni A , Otsuka T, and Ishihara M, Phys. Rev. C 52 (1995) R2334 25. Nakamura T. et al., Phys. Rev. Lett. 83 (1999) 1112
M E A S U R E M E N T S O F T H E n+p->d+y C R O S S SECTION F O R BIG B A N G NUCLEOSYNTHESIS WITH THE SPALLATION NEUTRON SOURCE AT THE LOS ALAMOS NEUTRON SCIENCE CENTER
E.-I. ESCH Los Alamos National Laboratory, LANSCE 3 ,MS H855 E-mail: [email protected] J. M. O'DONNELL, S. A. WENDER Los Alamos National Laboratory, LANSCE 3 D. BOWMAN, G. MORGAN Los Alamos National Laboratory,
P-23
J. L. MATTHEWS Department of Physics and Laboratory for Nuclear Science, MIT A large uncertainty in testing the elemental abundance predictions of Big Bang Nucleosynthesis models arises from the imperfect knowledge of the cross section for the formation of deuterium (D). An experiment to measure the n+H—»D+y reaction cross section is presently being performed at the high-energy neutron facility (WNR) at the Los Alamos Neutron Science Center (LANSCE) at Los Alamos National Laboratory (LANL). The white neutron source at WNR provides neutrons in the energy range from 10 keV to over 600 MeV. In the experiment we detect the y-rays emitted from the n+H—>D+y reaction with a highresolution Compton-shielded Germanium detector. The signature of the reaction is a y-ray whose energy depends on the incident neutron energy. The goal of the experiment is to measure the cross section in the neutron energy region between 30 and 500 keV to a precision of 3%. The setup and the current status of the experiment will be described.
1
Introduction
Big Bang Nucleosynthesis (BBN) models predict the dependences of the primeval abundance of certain light elements (D, 3He, 4He and 7Li), relative to H, on the baryon density. These elements are believed to have formed in the first 200 seconds after the big bang. The abundance of each isotope depends differently on the primordial baryon density Qb in the universe (see Figure 1); deuterium is the isotope with the most sensitive dependence and is thus termed the "baryometer of choice" for the early universe.
58
59
Qbhs 0.005
0.01
0.02
0.03
0.25 r o. 0.24
:
0.23
I
0.22
10 - 4
SH
qui
V
3 55 i—H
10~ 5 10" y
*
10" -10
?7 Figure 1: The primeval element abundance ratios.
In the early epoch of the universe, deuterium is in equilibrium with protons, neutrons and Y-rays- The deuteron is formed and then is photo-disintegrated, leaving the deuterium abundance small. Once the universe has cooled to a stage in which the average y-ray energy is sufficiently low, deuterons can form without dissociation. The n+p—>d+y process establishes the primeval deuterium abundance and its duration is limited by the neutron lifetime.
60
Since deuterium is processed through stars the measured local deuterium abundance is thought to be up to 50% lower than the primeval abundance. Buries and Tytler [ 1 ] have measured the deuterium abundance in high-red-shift intergalactic hydrogen clouds where it is believed that almost none of the deuterium has been processed by stars. The result quoted by them, with a 7% error, is D/H=(3.4±0.25)xl0~5. Buries et al. [2] have reviewed the astrophysical and nuclear input for the calculation of the baryon density and obtain a value of pb=(3.6±0.4)xl0"31 g cm"3. About half of the uncertainty in the above number comes from the uncertainty in the cross section for the n+p—>d+y reaction. Buries et al. stress the astrophysical importance of a more accurate measurement of the cross section for this reaction in the neutron kinetic energy range 30 to 600 keV. The cross section is well known at low energies where it is proportional to l/vn but cannot be extrapolated reliably into the energy region of interest for astrophysics due to the presence of the 3 S| resonance in the two-nucleon system.
2
The Experimental Setup
The high-energy neutron source at LANSCE (WNR) produces neutrons via spallation reactions generated by an 800 MeV pulsed proton beam. Since the neutrons have a continuous energy spectrum (see Figure 2), WNR provides a powerful tool for measuring cross sections over a range of neutron energies simultaneously. The spallation target consists of a tungsten rod 3 cm in diameter and 7.5 cm long which produces a yield of 5x108 neutrons per proton pulse. By surrounding the rod with 5 cm of beryllium to reflect and moderate the neutrons one can increase the low energy neutron flux. Monte Carlo calculations show that such a target can produce a factor of 4-7 more neutrons in the energy region of interest than a bare tungsten rod target. Figure 3 is a schematic illustration of the experimental setup. The experiment is situated on a beam line at an angle of 90° with respect to the incident proton beam. This beam line was chosen to enhance the low energy neutron flux relative to the high energy neutron flux, as seen by the Monte Carlo calculations shown in Figure 2. The neutron flight path from the spallation source to the proton target is 9.5 meters long. A neutron is captured by a proton in a polyethylene (CH2) or 6LiH target to form a deuteron, with the emission of a y-ray of energy equal to approximately one-half the kinetic energy of the incoming neutron plus the reaction Q-value (2.225 MeV).
61 The y-rays emitted by the target are then detected by a high-purity Germanium detector, with diameter 59 mm and thickness 75.5 mm. The Ge detector has a peakto-Compton ratio of 59:1 and is surrounded by an anti-Compton veto shield, which reduces the background produced by high energy neutrons capturing in the surrounding materials. Since the energy of the neutron-proton capture y-ray depends on the incident neutron energy, the energy dependence of this cross section can be measured. 1.00000
0.10000
tB 0.01000
^
0.00100
>0.00010
0.00001 10.0
1000.0
E„(MeV) Figure 2: Neutrons produced per proton. MCNPX calculation for a tungsten target at different angles.
The neutron flux is determined by a fission ionization chamber located approximately 3 m upstream of the sample. The energy of the incident neutron is obtained from its time of flight. With a proton pulse spacing of 4.2 |o.s it is possible to measure cross sections for neutron energies as low as 30 keV. With a data acquisition system that allows us to record time of flight, neutron flux, and y energy it is possible to extract the reaction cross section. For a given incident neutron energy, the capture y-ray energy increases from the Q-value of 2.225 MeV by En/2 , where En is the kinetic energy of the neutron. Figure 4 shows the capture y-ray energy as a function of neutron energy. Since the Ge detector has an energy resolution of 1.95 keV FWHM at 1.33 MeV, it should be possible to observe capture y-rays for neutrons with kinetic energies of 30 keV and above in 5-10 keV energy bins.
62
3
Status and expected Signal
In 2001 the experiment was set up and a short data acquisition run was performed. Figure 5 shows the y-ray spectrum measured during this run which lasted approximately 17 hours at a reduced proton beam current. The top curve in the upper figure is the y-ray spectrum in a 50 keV incident energy bin between 50 and 100 keV. The curve labeled "background" is the normalized background measured during the time interval before the high energy neutrons arrive at the target (see the lower figure). The curve labeled "subtraction" shows the spectrum in the 50-100 keV neutron bin after this normalized background has been subtracted. It is seen that the thermal neutron capture peak at 2.225 MeV is reduced by more than 70%. Due to the small cross section, low flux, and short running time it is not possible to extract from these data capture-y peaks for the neutron energy range of interest. In future runs, by optimizing the collimation of the neutron beam and running the proton beam at the normal rate of 100 Hz instead of the reduced rate of 20 Hz, one can achieve a higher neutron flux. A new anti-Compton shield design will allow the Ge detector to be placed closer to the target, thereby increasing the solid angle for y-ray detection. Taking into account these improvements plus the increased flux from the moderated spallation source, the expected count rate for the experiment may be estimated using the expression: A[ =
" e " / '°" Q AF N
—-D. F n
The terms are defined as follows: JVY: detected y-rays per second d2NiK,ulmJdQ.dE : neutrons per proton burst (8.4xl05 neutrons/(burst keV sr)) Q.T: solid angle subtended by the target (9.52x10"6 sr) AE„: neutron energy bin width (10 keV) Niare,: proton bursts per second (1.35xl04) dcr/dQ : cross section for n+p->d+y (9x10"30 cm2/sr) Q.D: solid angle of the Ge detector (0.6 sr) ED: efficiency of the Ge detector (0.06) 9target- areal density of the target (2.6x1022 atoms/cm2)
63
Pulsed Proton Beam
Fission Chamber
I
Collimation
Polyethylene Target 5x5 cm
Collimation
W Shield
W Target BGO Veto Shield
Flight Path: 9.5 m Figure 3: Schematic illustration of experimental setup.
2340 2320 • 2300
^
uT 2280 2260
^
_,.-•—
2240 —
2220
0
— -,-
50
•
—
—
- i
100
-
., .... ,, .
1
150
200
Figure 4: Emitted gamma ray energy versus kinetic energy of the neutron. Both energies are displayed in keV.
64
J
raw d a t a background subtraction
A, '—"WM
D
1500
>T_„.
2000
2500
3000
3500 E[keV]
Gamma Peak
—
G
to ->
o o
Intensity
1000
>
16000 14000
— —
10000
EE-
8000
~~
12000
— t i m e of flight Background intervall ^ = 5 0 - 1 OO k e V
eooo E~ 4000
E-
ot
2000
1 1
6000 700O TDC Bins Figure 5: y-ray energy and T0F spectra measured in the first run. The curve labeled "raw data" in the top figure is the spectrum in a 50 keV neutron energy bin; the curve labeled " background" is the normalized background. The spectrum labeled "subtraction" is the difference between the raw data and the background. The TOF interval in which the background was measured is shown in the bottom figure.
The numbers in parentheses are the expected improved parameters. Using these, one calculates a capture-y flux of 786 y's/day. With this count rate it should be possible to measure the cross section for the n+p—»d+y reaction to the desired precision of 3% within a few weeks of beam time. References 1. 2.
S. Buries and D. Tytler, Astrophys. J. 507, 732 (1998). S. Buries et al., Phys. Rev. Lett. 82, 4176 (1999).
PHONON PROPERTIES OF MATERIALS FROM NEUTRON RESONANCE DOPPLER BROADENING MEASUREMENTS J. ERIC LYNN Los Alamos National Laboratory, New Mexico, USA, 87545 At low temperatures the Doppler broadened widths of neutron resonances are strongly affected by the phonon characteristics of the material used for making the cross-section measurement. The Doppler width can be expressed in terms of the moments of the phonon spectrum carried by the atomic species with the resonant cross-section. Cross-section measurements made with tungsten and tantalum metals are reviewed here and compared with phonon information obtained by other methods. Applications of the method to a plutoniumgallium alloy and to some lanthanum barium cuprates are described briefly. We discuss possible extensions of the technique and how an epithermal flight path at the SNS may be advantageous.
1
Introduction
The most definitive method for determining the phonon frequency (or density-ofstates) spectrum of a crystalline material is by coherent inelastic scattering of very low energy (near-thermal) neutrons. In this method the phonon dispersion curves are obtained along the principal axes of the crystal lattice. These are fitted to an atomic force model, which is then used to determine the phonon frequency spectrum. The latter can also be obtained by incoherent inelastic scattering of lowenergy neutrons. Both methods have the disadvantage of being difficult or impossible to use if the thermal neutron cross-section is very high (which is common for materials containing heavy elements) or, in the case of the first method, a single crystal of adequate size cannot be prepared. With incoherent inelastic scattering only the overall phonon spectrum can be obtained rather than the spectra characteristic of the individual atomic species. Neutron resonance Doppler broadening spectroscopy offers a method of escaping those two difficulties, although the information obtained is much less detailed than the full spectrum shape, being limited to a few of the most important moments of the spectrum. The method involves the measurement of the crosssection as a function of neutron energy over suitable resonances that occur in the cross-sections of the nuclides that are constituents of the material under examination. From analysis of the detailed shape of these resonances simple moments or Boltzmann-weighted moments of the nuclide-specific phonon spectra can be obtained. In this paper I give a brief outline of the theory and experimental method, a brief review of experimental proof of principle, illustrations of applications to plutonium-gallium alloy and to some high-^c super-conducting
65
66
materials. Finally, I discuss the advantages that the SNS may bring to this field of application. 2
2.1
Theory and experimental requirements
Theory
Over a low energy neutron resonance the cross-section for a free nucleus at rest normally has the Breit-Wigner single-level form, which is a Lorentzian function of the neutron energy. We denote this by 0[(En). The effective cross-section that is measured for an assembly of atoms is aeffit=]dE'S(E')at(En-E') where S(E') is an energy transfer function (the Doppler broadening function) governing the transfer of energy with the environment of the nucleus in the neutron absorption process. The form of S was given originally for a gas by Bethe [1]. A little later Lamb [2] considered the nature of S for crystalline materials and concluded that for certain weak coupling conditions, which are usually met in the case of measurements of neutron resonance cross-sections, the Bethe gas formula is usually a quite good approximation provided the temperature is modified in a way that depends on the phonon properties of the crystal. In the opposite extreme of "strong coupling", the conditions for which do not seem to be attainable with neutrons but can be met with low energy gamma transitions (the Mossbauer effect), the predominant feature of the resonance cross-section is a "recoilless" term with the Breit-Wigner shape. Lamb's basic expression for the energy transfer function can be developed into a useful and accurate expression involving the moments of the phonon frequency function [3]: S(E') = (l/A^K)exp[-(E'-R)2/A2]{l+'Ln=~anHn[(E'-R)/A-\} Here, R is the recoil energy that would be carried off by the compound nucleus after absorption of a neutron by a free, stationary target nucleus, the H„ are the Hermite polynomials and A is the "Doppler width". For a gas sample, the last would have the form ^(2RkT), (T being the sample temperature and k Boltzmann's constant); for a crystalline material it is A = V(2/?(^v)T)
67
with (hv)j being the first of the Boltzmann-weighted odd moments of the frequency spectrum <(/!V)")T = Jdv {hv)n(v)co\\\{h\l2kT) Here, g(v) is the spectrum of phonon frequencies v carried by the resonant atomic species in the crystal. The expansion coefficients a„ are functions of the above odd moments and the simple (unweighted) even moments of the phonon frequency spectrum. The most important are a3 =
{(hvfy(l2A(hv)T)
a4 = ((hvf)T/(96R(hv)J2) These govern the magnitude of terms that give skewness and kurtosis to the Doppler-broadening function. At higher temperatures (kT considerably greater than (hv)) the odd Boltzmann-weighted moments can be expanded in terms of temperature and the simple even moments: ({hvf"+\
= 2kT({hv)2n) + {{hvfn+2)l6kT- ((/zv)2"+4)/[360(A7)2] ...
showing that the terms beyond the simple Gaussian then become small, and that there is convergence towards the simple gas model. At very low temperatures, of course, the odd Boltzmann-weighted moments approximate to the simple odd moments. It is thus apparent that the fitting of the effective cross-section profile over a range of temperatures from very low to moderate (say, room temperature) to determine the parameters of the Doppler-broadening function leads to important information on the major moments of the element- (or, more precisely, nuclide-) specific phonon frequency spectrum. 2.2
Experiment
The main experimental requirements are good resolution for the chosen resonances in the cross-sections of the material of interest. Usually the lowest energy resonances are the best choice because they are narrow and neutron time-of-flight spectrometer resolution rapidly worsens with increasing energy. High neutron fluence through the sample is required for statistical accuracy in determining the fine features of the Doppler-broadening function. In order to determine the first (simple) moment of the phonon spectrum measurements must be made at liquid helium temperatures. Liquid nitrogen temperatures are normally adequate to obtain complementary information on the second moment. At room temperature, or above, the gas model with estimated phonon corrections can be applied to measured data to obtain the nuclear resonance parameters, which can then be used to analyze the
68
lower temperature data for Doppler-broadening characteristics. Temperatures must be measured accurately. In the examples given below [4,5], the effective total crosssection was measured by transmission through samples with a range of thickness. The neutron source was the MLNSC spallation target at LANSCE. The proton pulse (FWHM) was 125 ns, and the flight path length was 58m. The neutron pulse is broadened by time-dispersion in the moderator-reflector assembly surrounding the tungsten spallation target. The neutron detector was a 6Li glass scintillator. Because of the very high intensity of neutrons arriving at this detector the neutron detection rate was measured in current mode rather than as discrete pulses, in order to avoid pulse pile-up problems. A severe problem with current mode detection is that longlife phosphorescence decay in the lithium glass is not eliminated and has to be determined and included in the data-fitting process. There are important phosphorescence modes in this glass with half-lives up to about 230 lis. 3
Measurements on Tungsten and Tantalum metals
Tungsten and tantalum metals (both bcc crystal lattices) have known phonon frequency spectra, which have been determined from coherent inelastic scattering experiments. They both have narrow low energy resonances in their cross-sections, and are therefore very suitable for testing the Doppler spectroscopy method. Examples of the data obtained [4] (the 7.6 eV resonance of tungsten at room temperature and very low temperature) and the fitted curves are shown in Figures 1 and 2. Relative count
1470
1490
1510
1530
1550
1670
Time (us) Figure 1. Neutron transmission as function of neutron time-of-flight around the 7.6eV resonance of tungsten. Sample thickness is 0.35mm and temperature is 31 OK
69 Relative count
1470
1490
1510
1530
1550
1570
Time (us) Figure 2. As Figure 1 except sample temperature is 15K.
Data over 6 resonances of tungsten (from 4.15eV to 47.8eV) have been analyzed to give a weighted mean value of the first moment (hv) = 20.2±0.4 meV. The value deduced from the published spectrum [6] is 20.0meV. The data on (hv)T up to room temperature can be fitted with a Debye model of the phonon spectrum with Debye temperature of 31514 K. This is very close to the value determined from specific data in a broad temperature range above 20K. The second moment has also been determined from the 15K data on the 4.15eV resonance as ((hv)2) = 0.04910.00lmeV2. This is to be compared with the value calculated from the published spectrum of 0.042meV2. Similarly good agreement is obtained for tantalum; our measured mean value for the first moment is 14.3+O.lmeV compared with 14.6meV deduced from ref.[7]. These results are a convincing demonstration of the validity and accuracy of the Doppler broadening method. 4
Plutonium- gallium alloy
An excellent example of application of Doppler-broadened neutron resonance spectroscopy is provided by the study of Pu-3.6at.%Ga alloy. Plutonium has a very complex phase diagram. The 8 (fee) phase is a low density phase (about 17% lower than the monoclinic oc-phase) and is favoured for important technological applications, but is not stable below about 583K. However, the addition of a few percent of gallium achieves the desired stabilization. The full thermodynamic
70
properties of this material are required for understanding its high pressure and temperature behaviour and the first few phonon moments are an important component of these properties. It has not yet been possible to prepare a suitable single crystal for coherent inelastic scattering measurements. Also, the thermal absorption cross-sections of the fissile isotopes are very high, which would add to the difficulty of making an inelastic scattering measurement. Our measurements using the Doppler method have been able to provide most of the required information [5]. The samples used comprised 93.9% 239Pu, 5.9% 240Pu and 0.1% or less of other isotopes. The resonance chosen for making the most accurate measurements is the 1.056eV resonance of 240Pu; its nuclear width is 32.7meV and its parameters have been very accurately measured at ORELA. The data at 15K on the thinnest sample (0.06mm) and its fit, giving a 1st moment of 8.22+0.12meV is shown in Figure 3. The fitted data at higher temperature can be represented by a Debye spectrum with 0 D = 127K. Many of the resonances at higher energies (especially those of 240Pu) have also been fitted and agree with this result although the accuracy of these measurements is somewhat lower. 5
4
O a> .>
4
CD
a. 1 o 3600
3800
4000
4200
4400
4600
Time-of-Flight (us) Figure 3. Transmission data across the 1 056eV resonance of 240Pu and fitted curve
The frequency spectrum carried by the gallium atoms in this alloy is also of great interest. The moments can in principle be determined from one of the gallium resonances at 96eV and llleV. The nuclear resonance parameters of these were determined from samples of gallium metal, and also, by using cooled samples, the phonon moments of gallium metal in its orthorhombic a-phase, obtaining (hv) = 17.4±0.4meV and 0 D = 270±6K. In the Pu-Ga alloy, unfortunately, both resonances lie extremely close to resonances of 239Pu. It is still possible to obtain information from the l l l e V resonance because the neighbouring Pu resonance at 110.5eV is
71
very weak (although still of similar magnitude to the very weak dip due to the gallium resonance in the transmission curve). The final value for the mean phonon energy of gallium in the 8-phase alloy is 16.3±1.4meV, almost twice the mean phonon energy carried by the plutonium atoms. Recently, an experiment was started with Pu-Ga alloy samples in which the main isotope 239Pu is replaced by 242Pu, which has far fewer resonances, none close to the gallium resonances. This will allow a much more accurate determination of the gallium frequency and may allow us to determine if the gallium atoms are vibrating as a coherent part of the overall lattice or are vibrating in a simple local mode. Another application has been to the lanthanum barium cuprate, La2.;tBa^Cu04, high temperature superconducting system. While there is much evidence that the superconducting mechanism is connected mainly with spin fluctuations in the copper oxide planes rather than the electron-phonon coupling of conventional superconductors, there is also considerable evidence that phonons are involved. We have therefore begun experiments to see how the mean phonon energy on each of the metallic species varies with the Ba fraction x, on which the critical temperature depends in a complicated fashion. So far we have examined the cases x =0, 0.125 (non-superconductimg) and x = 0.15 (superconducting). There appear to be strong variations of the mean phonon energy amongst atomic species and as a function of x; effective Debye temperatures vary by over 100K with changes in the latter. The most accurate results are obtained for lanthanum because of its strong 72eV resonance; results are 0 D = 254±2meV, 187±3meV and 248±2meV at x =0, 0.125, 0.15, respectively. Copper is difficult, the only usable resonance being at 230eV where resolution is rather poor. Debye temperatures are 424±16, 545±15 and 498±12meV for the above values of x. Barium results (from the 82eV resonance) are 0 D = 302+9 (*=0.125) and 0 D =182±6 meV (x = 0.15). 5
Doppler Resonance Spectroscopy at the SNS
It appears that it will not be possible to have an epithermal flight path much longer than 100m at the SNS. This will permit resolution a little better than that of the 60m flight path at LANSCE in the low energy region in which proton pulse width is dominated by moderator time dispersion. At higher energy the resolution of the existing ORELA facility is far better. The resolution of a 100m flight path compared with the Doppler width, at very low temperature, for samples of various mass number and Debye temperature combinations is shown in Figure 4. It is apparent that resolution is sufficient to make useful measurements on resonances up to 200 to 300eV but not beyond.
72 Energy resn. or width (eV)
1
10
100
1000
Neutron energy (eV) Figure 4. Resolution (FWHM) at 100m flight path on SNS ( . .) compared with Doppler width in following cases: 1) A = 240, 0 D = 200K (+ + +), 2) A = 120, 6 D = 300K (* * *), 3) A = 240, 0 D = 350K (ODD)
The main advantage of the SNS will be its much higher neutron intensity. This will enable measurements to be made on much smaller samples than hitherto. It will allow measurements to be made of capture cross-sections using very thin samples rather than by transmission; thus phosphorescence in the detector, a major problem in analyzing data from current mode collection, will be avoided. It will also allow the use of very weak resonances (including p-wave resonances), thus extending the range of elements accessible to the technique. Acknowledgements I wish to extend my acknowledgements to my colleagues in this project: Walter J. Trela, George H. Kwei, Kai Meggars, Vincent W. Yuan and Luke Daemen, and also to Steve Sterbenz and Bard Bennett for helpful discussion and support. This work was done at Los Alamos under DOE auspices (contract W-7405-ENG-36). References
1. 2. 3. 4. 5. 6. 7.
H.A.Bethe, Rev.Mod.Phys. 9, 140 (1937) VJ.E.Lamb,Phys.Rev .55,190 (1939) J.E.Lynn and W.J.Trela, Nucl.Instr.Meth.Phys.Res.B, 108,147 (1996) J.E.Lynn , W.J.Trela and K.Meggars, Nucl.Instr.Meth.Phys.Res.B (in press) J.E.Lynn, G.H.Kwei, W.J.Trela, V.W.Yuan et al,Phys.Rev.B58,l 1408(1998) S.H.Chen and B.N.Brockhouse, Solid State Comm. 2, 73 (1964) A.D.P.Woods, Phys.Rev. 136, A781 (1964)
APPLIED NUCLEAR PHYSICS AT SPALLATION NEUTRON SOURCES
PAVEL OBLOZINSKY National Nuclear Data Center, Brookhaven National Laboratory, E-mail: oblozinsky&.bnl.Kov
Upton, NY 11987, USA
Research opportunities offered by spallation neutron sources for applied nuclear physics are reviewed, in particular having in mind the SNS machine under construction at Oak Ridge. Discussion is focused on nuclear data needs for 4 groups of applications. These include nuclear power (energy production, transmutation), defense (criticality safety, radiochemical diagnostics), and other applications (material analysis, medical, standards and resonance parameters), as well as needs for nuclear reaction model calculations (level densities, photon and neutron strength functions). It is concluded that the new SNS machine represents an important tool to meet future nuclear data needs in the U.S.
1
Introduction
In general, applied nuclear physics is viewed as a topic where nuclear data play an important role. Our approach in the present review paper will follow this established wisdom. Thus, we will first attempt to identify future nuclear data in various applications and then indicate how these needs can be met by using spallation neutron sources. In doing so we will naturally focus on the Oak Ridge SNS machine. The Oak Ridge SNS machine [1], to be commissioned in 2006, has several design features that will provide framework to our discussion. Most importantly, it is high flux machine based on a 2 MW proton beam producing 2.08xl014 protons per pulse. Considering repetition rate of 60Hz and an average of 30 neutrons per incident proton, the SNS will produce as much as 3.7x1017 neutrons per second. This impressive parameter is adversely influenced by relatively modest time resolution of about 350 ns FWHM, compared, say, to 7 ns of the n-TOF machine at CERN. Considering furthermore relatively short flight path of 20 meters, one can conclude that, due to modest energy resolution, a practical upper energy limit for cross section measurements is about 100 keV. Therefore, the present paper is focusing on nuclear data needs and applications in this low neutron energy range. On the other hand, it should be stressed that due to high flux the Oak Ridge SNS machine offers several unique possibilities for neutron cross section measurements. Thus, one can employ small targets and in particular radioactive targets, measure small cross sections, as well as perform high precision cross section measurements. The paper is organized as follows. First, we list relevant nuclear physics applications, then proceed by discussion of nuclear reaction data needs for each of
73
74
them. Finally, we draw conclusions on the relevance of the SNS machine for measurements addressing these needs. 2
Applied Nuclear Physics
We start our discussion by providing a list of relevant topics in applied nuclear physics. Our list includes 10 items organized into 4 groups of applications, nuclear energy, defense, and other applications, and also reaction model applications: • • • •
Nuclear power applications (energy production, transmutation), Defense applications (criticality safety, radiochemical diagnostics), Other applications (material analysis, medical, standards, and resonance parameters), and Reaction model applications (level densities, strength functions).
This list is primarily deduced from 4 recent nuclear data conferences and meetings: •
•
•
•
3
3.1
International Conference on Nuclear Data for Science and Technology, October 2001, Tsukuba, Japan [2]. This type of conferences is held triennially and it represents a major event for nuclear data community worldwide. IAEA Advisory Group Meeting on Long Term Needs for Nuclear Data Development, December 2000, Vienna, Austria [3,4]. This meeting addressed needs of many applications and formulated priorities for future nuclear data development work. Meeting of the U.S. Cross Section Evaluation Working Group, CSEWG, November 2001, Brookhaven, USA [5]. The CSEWG organization, in place since 1966, is responsible for developing the U.S. evaluated nuclear reaction data library ENDF/B. The core of the ENDF/B library consists of cross sections for 325 nuclei of practical importance in the incident neutron energy range from 10"5eV to 20 MeV. NEA High Priority Request List [6]. This list of requested nuclear data, maintained by NEA Data Bank, Paris, was recently reviewed at the NEA WPEC meeting, April 2001, Santa Fe, USA. Nuclear Power Applications
Energy Production
The nuclear power industry is traditionally the main customer of evaluated nuclear reaction data. Since the last US nuclear power plant (NPP) was commissioned some
75
20 years ago, this healthy relationship between nuclear data community and nuclear industry has slowly eroded. Notably, however, there is recent shift in the attitude towards nuclear energy in the U.S. that can be characterized as "revival of nuclear energy" or "nuclear renaissance" [7]. This observation is supported by the fact that DOE is involved in 3 major activities in support of nuclear power [8]: • •
•
Nuclear Power Plant Lifetime Extension. The process is underway to extend licensing of 103 US nuclear power plants for another 20 years. Nuclear Power 2010. The goal is to have new order/orders for commercial NPP in place by 2005, with deployment by 2010. The project mainly addresses regulatory procedures and issues. New Generation of Reactors, Gen-IV. Main criteria for the new generation of reactors include economic competitiveness, enhanced safety & reliability, and enhanced resistance to proliferation risks. Four classes of reactors are under consideration, water cooled, gas cooled, liquid metal cooled and innovative. The current effort is focused on a roadmap study to select 6-8 concepts including R&D needs. The roadmap report should be completed in fall 2002. Then, R&D should follow, target for deployment being 2030.
It should be noted that one observes a tendency towards revival of nuclear energy in other countries as well. Thus, the International Atomic Energy Agency (IAEA) launched an International Project on Innovative Nuclear Reactors and Fuel Cycles, INPRO. The ongoing phase I of the project, scheduled for 2001—2003, should identify user needs and selection criteria for future generation of reactors. As for data needs for nuclear power applications, one can make several general observations. Current evaluated nuclear reaction data libraries are fairly good for "classical" reactor concepts, where extensive benchmarking has been performed. For "non-classical" concepts, an improved library is needed. In particular, the nuclear industry is not interested in individual improvements of evaluations of specific nuclides. Rather, the industry needs a complete library addressing new needs that would fall into 2 categories. First, data uncertainties (co-variances) should be provided for all evaluations. This would allow more realistic treatment of design and operational margins, and lead to significant reduction of constructional and operational costs. Second, improved evaluations are needed for new coolants, absorbers, structural materials and fuel cycles. It should be kept in mind that even though many of these materials are not entirely new, the neutron energy range of new reactors would in principle differ from that of importance for classical reactors. Detailed discussion of some of the above points can be found at the keynote paper of Max Salvatores at the Tsukuba Conference in 2001 [9], see also several other papers [10]. Below we provide just a couple of examples for illustration: •
Innovative fuels. Matrix diluents for composite fuels would require new materials such as Zr, Mg, Y, Ti, W, and also Ce, V, Mo and Al. Capture, elastic
76
and inelastic cross sections are needed with improved accuracy. Further data needs are related to advanced fuel design including high enrichment and new composition. New coolants. Of particular interest are Pb and Pb/Bi, with data needs as above. Extended burn-ups and homogenous recycling of minor actinides. Needed are neutron interaction data on minor actinides and also decay heat data. Th fuel cycle. Improved data are needed for 232Th(n,y) as well as data on 31Pa,
• • •
232,233, ,
The last point is of specific interest. To this end, IAEA addressed needs of the Th fuel cycle in 1999 [11]. It was concluded that data for 7 nuclei, namely, 232Th 231,233 Pa and 232 . 233 . 234 . 236 TJ should be improved. In particular, capture and fission for thermal and fast reactors are needed with accuracy summarized in Table 1. Table 1. Required precision (%) of capture and fission cross sections in thermal and fast neutron energy range for 7 nuclides involved in the Th-U fuel cycle [11]. 2J2
Capture thermal Fission thermal Capture fast Fission fast
3.2
Th
1 1-2 5
i3,
Pa 10 10 20
233
Pa
5 3-10 20
'm\J 10 10 50 20
2
-"u
5 1 3 1
234
3 5 3
U
m
U 20 10 5
Transmutation
Disposal of radioactive nuclear waste represents one of the most important challenges for nuclear power industry. More generally, this can be viewed as a part of a broader problem of the back-end of the fuel cycle that involves two issues. The first issue is Pu management. To this end, one considers various strategies, such as recycling, consuming and immobilization, with improved capture data needed in epithermal resonance region for 238.240-242pu The second issue is actual transmutation. Of particular concern is transmutation of Pu and minor actinides, where transmutation means fission. In addition, one should take care of many long-lived fission products, where transmutation means capture. It is instructive to analyze impact of cross section uncertainties on transmutation process. As an example, we show a case discussed by Salvatores [9]. He considered minor actinides (MA) and their transmutation in a reactor with dedicated core. Fuel was a mixture of Pu and minor actinides, with TiN layers and ratio of Pu/MA = 0.6. Cross sections uncertainties were treated realistically as 5— 30% depending on isotope, reaction and energy range Then, calculated contributions to uncertainty in criticality, Ak/k, were found to be substantial as
77 shown in Table 2. It is seen that the current cross section uncertainties are far too high, leading to unacceptably high total uncertainty of 2.13% in this specific case. Table 2. Impact of cross section uncertainties of 6 minor actinides on criticality, Ak/k transmutation scenario using reactor with a dedicated core [9].
Isotope 238 Pu MI Am 242m Am MJ Am w Am Z45 Am Total
Capture 0.17 0.99 0.01 0.54 0.15 0.03 1.16
Fission 0.97 0.62 0.53 0.26 0.46 1.14 1.78
(%), in
Total Isotope 0.99 1.17 0.53 0.60 0.48 1.14 2.13
Data needs for transmutation include 3 categories of nuclei: •
•
•
4
4.1
Minor actinides. It should be noted that fast neutron spectrum gives best conditions for transmutation. Needed are many neutron interaction data. Target accuracies can be found in the NEA High Priority Request List [6]. Lanthanides. They are of substantial practical importance in view of difficulties to separate lanthanides chemically from minor actinides. Again, fast spectrum gives best condition for transmutation. Fission products. Of concern are long-lived fission products, such as 99 Tc, 129I and 135 Cs. Needed are improved capture data, for more details see also [12].
Defense Applications
Criticality Safety
The basis of the ongoing US criticality safety effort is the US Defense Nuclear Facility Safety Recommendation, namely, the well known recommendation 93-2. In response to this recommendation, DOE Environmental Management sponsors the Nuclear Criticality Safety Program. An inherent part of this program is nuclear data measurement, evaluation and validation, with contributions mostly from ORNL (see Valentine [13] and Leal [14]), LANL and ANL. Of primary importance are measurements at the Oak Ridge high resolution ORELA machine in the resolved resonance and unresolved resonance region. So far, neutron total and/or capture cross sections have been measured for .
19
F, 2 7 Al, 2 8 W 0 Si, 3 5 ' 3 7 Cl, 2 3 3 ' 2 3 5 U,and
78
•
,6
0, 39 ' 41 K, 55Mn, 238U are under completion.
Future nuclear data needs include improved fission product cross sections for burnup credit. This is related to regulatory procedures for storage and transportation of nuclear materials that would allow accounting for reduced reactivity due to neutron interaction with fission products in irradiated fuel. Another motivation is related to much longer fuel utilization time, and thus much higher concentration of fission products in spent fuel as foreseen in future. Additional requirements are expected from new generation of reactors, in particular new absorbers, coolants, innovative fuels, fuel cycles and new neutron energy range. 4.2
Radiochemical Diagnostics
Radiochemical diagnostics is a topic of high importance for analysis of archived nuclear test data. A key challenge in applying this technique represents analysis of data from radiochemical detectors that are activation in nature. There are about 60 elements with properties that would make them good candidates for these detectors, including • • •
Ti, V, Cr, Mn, Fe, Br, Kr, Y, I, and Eu, Sm, Gd and Tm.
To make use of these activation detectors, precise cross sections are needed for (n,2n), (n,n') and (n,y) for the whole network of reactions. In addition, needed are not only reactions on ground states but also for production and depletion of isomers. 1 t
4
I +••
-
I
•
{
»
?
c
Y ' *
•
•
*
•
it,1
0.01
O.t
E, (M«V)
Figure 1. Comparison of experimental (points) and theoretical/evaluated (full curve) values for the ratio of 151Eu(n,y)/153Eu(n,y) cross sections [15].
79
The overall procedure for radiochemical diagnostics is as follows. Experimentally, one loads detector material into a device, conducts test and measures activation products. Then, one analyzes measured data. To this end, one needs a set of evaluated energy dependent cross sections that couple all species and states (detector set). With this set one performs reaction network calculations using assumed neutron flux, calculates activation products and forms their ratios. These predictions are compared with measured values so that neutron flux can be deduced. Evaluations of cross sections for detectors in the rare earth region, Gd, Eu and Sm, are based on careful reaction model calculations as discussed by Hoffman at the recent CSEWG meeting [15]. As an example we show in Fig. 1 evaluated and experimental data for the ratio of l51Eu(n,y)/,53Eu(n,y) cross sections. It is seen that experimental data are not mutually consistent and much better measurements are needed. Precise new measurements of this type are required in the energy range below 0.1 MeV for many isotopes of 60 elements as indicated above. 5
5.1
Other Applications
Material A nalysis
Several promising nuclear technologies are emerging in the field of material analysis. Use of cold neutrons is one of them. Another very promising technique is the resonance neutron capture in the neutron energy range of about 1—10 keV. This new technique is termed Neutron Resonance Capture Analysis (NRCA) and it is based on a pioneering paper of Postma et al. [16]. One needs pulsed epithermal beam and time-of-flight system associated with the measurement of neutron capture. Composition of materials is deduced from observed neutron resonances. This technique has extremely high selectivity, many elements can be recognized quickly without complicated analysis. The technique is non-destructive. The irradiation time and related undesired activation of samples are very low. The SNS machine seems to be well suited for NRCA measurements. 5.2
Medical Applications
Medical applications use both diagnostic and therapeutic radio-nuclides that are produced either by reactors or cyclotrons. Mainly p + emitters are used in diagnostics while P" emitters (as well as other radiation sources) are used in therapeutics. Thus, of primary interest for high flux neutron machines would be production of neutron rich therapeutic radio-nuclides. To this end, the following studies are of interest [17]:
80
• • •
5.3
Improved production and expansion of the current list, including Cu, Cu, Sr, ,53 Sm, ,59 Hoand ,66 Ho. Study of low energy neutron reactions, including fission, capture, and double capture such as 164Dy(n,y) 165Dy(n,y) 166Dy -> 166Ho. Exploration of alternative production ways via (n,p) reactions with fast neutrons from breakup and spallation neutron sources. 89
Cross Section Standards
Internationally recognized set of neutron cross section standards, as included in the ENDF/B library, consists of 9 reactions. Namely, H(n,n), 3He(n,p), 6Li(n,t), l0 B(n,a), I0B(n,ay), C(n,n), 197Au(n,y), 235U(n,f) and 238U(n,f). These standards should be regularly reviewed and improved. For instance, reactions on boron, 10B(n,a) and l0B(n,ocy) serve as standards in the neutron energy range from thermal to 250 keV. In order to reach required precision of ~ 1 % the following cross sections would need better measurement [18] that would advantageously use a high flux machine: • 10B(n,ocy), and • 10B(n,tot), including its components 10B(n,Oo), 10B(n,a,) and 10B(n,n). 5.4
Resonance parameters
Resonance parameters of radioactive nuclei are virtually unavailable. Well known compilation by Mughabghab [19] includes resonance parameters for 280 stable nuclei from 24Na to 253Cf. These parameters are of critical importance for cross section evaluations included in the general purpose file, ENDF/B, that serves many practical applications. There are substantial data needs for resonance parameters, for example, for 134'135Cs, 153Gd, and for many other radioactive nuclei. 6
Reaction Model Applications
Nuclear reaction model calculations represent an indispensable tool in evaluation of nuclear reaction data. These calculations rely on precise input parameters that are often very difficult to obtain. Two key parameters in statistical reaction model calculations are nuclear level density and optical model potential. In both instances, SNS machines can be most usefully employed to produce requested parameters, see also discussion by Herman [20]. 6.1
Level Densities
One of few possibilities to obtain direct experimental information on level densities is provided by neutron resonances. At the neutron biding energy, level densities for
81 spin J = 0 are equal to 1/D0, where D0 is the average s-wave neutron resonance spacing. Values of D0and also Dj are deduced from observed neutron resonances. Spallation neutron sources offer substantial possibilities for measurements of resonance parameters and resonance spacings of radioactive samples for which there are, at present, very little data. 6.2
Strength Functions
Photon strength functions are important for statistical model calculations of neutron capture and also to account for gamma competition in particle reaction channels. Of particular interest for obtaining photon strength functions would be measurements of y spectra using the technique of average resonance capture. Needs are primarily in the area of radioactive nuclei and also in Ml and E2 photon strength functions. Neutron strength functions, S0 and Sj, define neutron absorption cross sections at low energies. Values of So and Si are thus of critical importance for optical model parameterization at these energies. Neutron strength functions are derived from average resonance parameters. Again, needs are primarily for radioactive nuclei. 7
Conclusions
In the present paper, we discussed applied nuclear physics at spallation neutron sources. Considering in particular the SNS machine at Oak Ridge the following conclusions can be drawn: • There is a strong case for the epithermal beam line. • Particularly attractive are measurements using small/radioactive samples, measurements of small cross sections and precise (reduced uncertainties) measurements. • Dominant reaction type to be measured is capture. This means that often there will be an overlap of interests with nuclear astrophysics. • High flux SNS machine should be viewed as a complementary to the high resolution ORELA machine. • Cross section measurements at SNS would be of vital importance for future US Nuclear Data Program. Acknowledgements The author is grateful to numerous colleagues for enlightening discussions. He would like to thank Mark Chadwick and Bob Haight (both LANL), Charlie Dunford and Said Mughabghab (BNL), Rob Hoffman and Frank Dietrich (LLNL), Ed Cheng (TSI), Allan Carlson (NIST), Don Smith and Dick McKnight (ANL), Paul Koehler
82
and Tim Valentine (ORNL), Gabor Molnar and Tamas Belgya (Budapest), as well as Mike Herman (IAEA). References 1. Koehler P.E., Comparison of white neutron sources for nuclear astrophysics experiments using very small samples, Nucl. Inst. Meth. Phys. Res. A460 (2001) pp. 352-361. 2. Proceedings of the International Conference on Nuclear Data for Science and Technology, October 2001, Tsukuba, Japan, to be published in J. Nucl. Sci. Tech. 3. Eds. Muir D.W. and Herman M., Long Term Needs for Nuclear Data Development-Summary, Report INDC(NDS)-423 (IAEA, Vienna, May 2001). 4. Ed. M. Herman, Long Term Needs for Nuclear Data Development--Text of Papers, Report INDC(NDS)-428 (IAEA, Vienna, December 2001). 5. Summary of the 51s' Cross Section Evaluation Working Group Meeting, November 6-7, 2001, Brookhaven (Report BNL, 2001). 6. The NEA High Priority Request List, www.nea.fr/html/dbdata/hprl/. 7. Marcus G. H. and Levin A. E., New Designs for the Nuclear Renaissance, Physics Today, April 2002, pp.54--60. 8. DOE Office of Nuclear Energy, Science & Technology, www.ne.doe.gov. 9. Salvatores M., Future Nuclear Power Systems and Nuclear Data Needs, see Ref. 2, to be published. 10. Eds. Oblozinsky P. and Gandini A., Nuclear Reaction Data and Nuclear Reactors: Physics, Design and Safety (World Scientific, Singapore, 1999). 11. Ed. Pronyaev V., Summary of the Consultants' Meeting on Assessment of Nuclear Data Needs for Thorium and other Advanced Cycles, Report INDC(NDS)-408 (IAEA, Vienna, August 1999). 12. Harada H., contribution to the present Workshop. 13. Valentine T., contribution to the present Workshop. 14. Leal L., contribution to the present Workshop. 15. Hoffman R.D., Rare Earth Nuclear Reaction Cross Sections for Radiochemical Diagnostics, see Ref. 5, Attachment 5-6, pp. 1-18. 16. Postma H. et al., Neutron-resonance capture analysis of materials, J. Radioanal. Nucl. Chem. 248 (2001) pp. 115-120, 17. Qaim S.M., see Ref. 4, pp. 7—16. 18. Carlson A., Requirements of the Neutron Cross Section Standards, in Report INDC(NDS)-425 (IAEA, Vienna, June 2001), p. 23. 19. Mughabghab S.F. et al, Neutron Cross Sections: Neutron Resonance Parameters and Thermal Cross Sections (Academic Press, New York 1981 and 1984). 20. Herman M., contribution to the present Workshop.
APPLIED PHYSICS MEASUREMENTS AT THE CERN N_TOF" FACILITY E. GONZALEZ on behalf of the n_TOF COLLABORATION. CIEMAT, Avda. Complutense 22, Madrid-SPAIN enriaue. eonzalezOiciemai. es The present experimental program of the recently constructed neutron time-of-flight installation at CERN, n_TOF, is dominated by the objectives of a shared cost action project of the EU for the measurement of neutron cross section needed for ADS development and nuclear waste transmutation, nTOF-ADS. This paper presents the motivations and the list of reactions and isotopes considered in this part of the n_TOF program. The experimental methods and the present measurement schedule are also briefly discussed.
1
Introduction and motivations for the measurements program
Nuclear waste is one the main problems for the public perception of the nuclear energy production and for the sustainability of this energy source and, although deep underground repository seems to be a scientifically proven and technologically viable solution for the nuclear waste for the first thousands of years, this option presents difficulties for social acceptability. For this reason, nuclear waste transmutation has been proposed as a way to substantially reduce (1/100) the inventory of the long lived component of the nuclear waste, mainly the transuranium actinides. Actinide transmutation is proposed to take place by fission in nuclear systems, critical reactors or subcritical Accelerator Driven Systems (ADS), in most cases of fast neutron energy spectra and using specific fuel compositions, much richer on high mass trans-uranium actinides. In addition, the transmutation of Long Lived Fission Fragments, LLFF has also been proposed using neutron absorption (mainly by radioactive capture) normally in the thermal and epithermal neutron energy spectra. The presently available neutron cross sections have been mainly motivated by the exploitation of the U-Pu cycle in nuclear reactors of neutron thermal spectrum and the design and operation of experimental fast U-Pu nuclear reactors. As a consequence, even when the present nuclear data base is sufficient for the conceptual design of the transmutation oriented nuclear devices, critical reactors or ADS, and for the first order evaluation of the impact of the transmutation technology in the nuclear waste management, nevertheless the detailed engineering designs, safety evaluations and precise performance assessment of critical reactors 1
Project financed by the European Commission shared cost action FIKW-CT-200000107
83
84
and even more ADS, operating for transmutation with fuels including large fractions of transuranic isotopes, require more precise and complete basic nuclear data. New isotopes, new energy regions and new reactions play a major role on the behavior of those transmutation nuclear systems and the corresponding cross section data are needed with better quality. Figure 1 shows a comparison of the composition of fuels of different nuclear plants including the typical PWR fresh and irradiated fuels, and the fuel for the typical transmutation devices of two scenarios for the fuel cycle, the equilibrium cycle in a version of the double strata and a scenario for the phase-out of the electricity production by nuclear fission. The new fuel compositions modify severely the role of the different isotopes for the global operation of the reactor and in particular for its performance in transmutation terms, as can be observed in Figure 2 and Figure 3 that show the relative contribution to capture and fission from different isotopes in the fuels previously described. i
H oi ;
i
'
r? "
%
£
iri
1
^ 0 Fresh PWR f DPWR50GWd/THM ' S Trans. Equilibrium MOX+ADS B Trans. ADS Phase-OUT
S V
•
41 5
—-*
j
s
t
R i°
^* £
S c 6? 3
i
5 i -•
* ^
3
I
co CO
# i 3 i5
0.10% —
•
; !
(£5
1 c c
as CO
f
s!
ft
EL. J i
0 U
o o —
a? p
i
5
c
-
II : o
c
c
: *
> :
iS
f '
a9
,
.
,
rl
c
r
i
J
1 1°°*,^
1 m—S
C
i
F if
Figure 1: Comparison of the composition of fuels of different nuclear plants: PWR (fresh and irradiated fuels) and for two Transmutation devices.
85
70%
60%
0 PWR Fresh fuel. Capture D PWR 50 GWd/THM. Capture BTrans. Equilibrium MOX+ADS. Capture
50%
B Trans. ADS Phase-OUT. Capture
30%
Figure 2: Contribution to capture from different isotopes of the fuels of different nuclear plants: PWR (fresh and irradiated fuels) and for two Transmutation devices. 100% 90%
0PWR Fresh fuel. Fission
80%
D PWR 50 GWd/THM. Fission
70%
H Trans. Equilibrium MOX+ADS. Fission H Trans. ADS Phase-OUT. Fission
60% 50% 40% 30% 20% 10% 0%
sft5 oj^tN c^o
a^S SX: e>^
ass* Sbo coo
sWss CCNO doo
J3L
Figure 3: Contribution to fission from different isotopes of the fuels of different nuclear plants: PWR (fresh and irradiated fuels) and for two Transmutation devices.
86
One general topic considered in the nTOF program is the thorium fuels. The main reactions on these fuels are the neutron capture by 232Th that, after fast radioactive decays breeds U, and the fission of the U that finally provides the neutron multiplication. 232Th has a smaller nuclear mass than 238U, as a consequence its neutron irradiation produce mainly uranium isotopes and much less transuranium actinides (protactinium is also produced). This allows to generate a much better nuclear waste in nuclear fuels with 232Th matrix than in 238U matrix fuels. For this reason, nuclear fuels with 232Th matrix had been proposed both: for the transmutation devices for the present nuclear wastes, and for a new generation of "cleaner" nuclear energy production systems with a much smaller generation of trans-uranium actinides in the irradiated fuel. In both cases, fast and thermal neutron spectra had been proposed. The list of new isotopes that play an important role in the thorium fuel cycle includes: 230,232Th, 232.233.234.236u and 23'.232-233pa. Some of the important reactions for this cycle are: 233 M3 m U + n->2^U_... (7 neutrons before Pu) "Th + n > Th-> Pa->™U, 233 212 U + « -> Thermal and Fast fission, Th + n Fast fission 232 233 U+n ->2n+232U Th + n-+2n + 23,Th->23,Pa; 23,Pa + n ->"2pa^232U; 100% -p 90% 80% 70%
•
•
i
- Pu-240 - Pu-241 - Pu-238
90% -
..)
80% 70% •
:
60%
1
50% 40%
100% •
i F
"Iff Capture
ill-
H*
30%
m
(if20%
!"]
100
10J
* * »* * . . * . * > '
— — — — —
Pu-239 Pu-240 Pu-241 Pu-238 Am-241
-
,
. ..*
.-..
~ .
> 195SF
flit \
J
r
50% • 40% " 30% • 20% "
1 10
i „
60% -
i
1
,
a
-!-i 10% •
1
10" 105 10" 10' Neutron Energy (eV)
\ 1
"t=Ssi 10" 10° 10° 10' Neutron Energy (eV)
Figure 4: Capture and fission energy range of interest in transmutation devices (Th-based fast ADS). To reach 99% of the full capture rate neutron energy has to be integrated up to 1 MeV. In the case of fission the 99% require to extend the energy range to nearly 10 MeV.
The second generic field of nuclear data needs is related to the use of ADS for nuclear waste transmutation. Effective nuclear waste transmutation requires the use of special fuels with reduced intrinsic safety features (lower (3efr, lower Doppler
87
effect, ...) making more difficult to operate in critical condition. In addition new chemical forms of the proposed fuels and the need for a much extended fuel burnups require larger operational flexibility than what is available in critical reactors. The subcriticality of the ADS and its external intense neutron source provides the required flexibility without loss of safety margins. However, to operate a subcritical device requires a neutron spallation source that brings even faster neutrons (En>5MeV) into the system (but only few per cent) and consequently new reactions such as (n,xn) become more important. In addition, the fact that lead and Pb/Bi alloys have often an important role in ADS, both as the target of the spallation source and as coolant of the nuclear core, make specially relevant the cross sections of these elements. The improvements required can be classified on the following topics: • Measurements of the basic cross sections (elastic, capture, fission, total and inelastic) on many high mass transuranic isotopes for which there is none or just one single experimental data set, but that appear in sizeable amounts in most of the transmutation schemes, based on multiple recycling of minor actinides or transuranics. The most important energy range covers from thermal to about 1 MeV for capture and thermal to several (20) MeV for fission, Figure 4. • Measurements of the basic cross sections of medium and long lived fission and spallation products. For most of these isotopes there are very few measurements and the available data are uncertain. Improvements are required to evaluate the possibility of their transmutation in dedicated devices. The most important energy range for these measurements is again from thermal to several (20) MeV. • Measurements of fast neutron reactions, in the energy range from few keV to 20 MeV (capture, fission, elastic, inelastic, (n,xn), (n,a), etc), with usual actinides, fission products, coolant and structural materials, Figure 4. Experimental data are scarce or not available at all in many relevant isotopes. • Evaluation of available experimental data to compute cross sections and dissemination of the evaluated cross sections results through the international agencies co-ordinating the distribution of these data. • Measurements of higher energy (10 MeV to several hundred MeV) neutron cross sections and improvement and evaluation of nuclear models for the processes appearing during the operation of spallation neutron sources, in the interactions of neutron and charged particles of these energies. As a summary of these needs the wish list of isotopes and reactions for nuclear Tin
919
911
919
911
waste transmutation should include, as actinides: Th, " Th, Pa, Pa, Pa, 232 U, 233U, 234U, 235U, 236U, 238U, 237Np, 238Pu, 239Pu, 240Pu, ^'Pu, 242Pu, 244Pu, 241 Am,242mAm,243Am,242'243-244'245'246-247'248Cm; as long lived fission fragments: 99Tc, 129 I, 15,Sm, 79Se, 126Sn, 135Cs, 93Zr, 107Pd; and as new coolants and structural materials: 2 0 < ™ 8 Pb, 2 0 9 Bi.
88
On the other hand the main reactions are: Capture, that equilibrates fission in the neutron multiplication by parasite absorption of neutrons and generates the actinide nuclear wastes but at the same time breeds fissile isotopes and is the intermediate step to the transmutation fission for some actinides and the main transmutation reaction for LLFF; Fission, that produce the neutron multiplication and the reactor energy and is the final actinide transmutation reaction; Inelastic scattering, that shapes the neutron spectra in fast reactor cores (coolant, fuel and structural materials); and (n,xn), that contribute to the shaping of the neutron spectrum and to the neutron multiplication near the spallation target but also generates the most dangerous nuclear waste in the Thorium cycle. The largest activities in these directions, inside Europe, are co-ordinated within the n TOF and HINDAS projects, both including experimental measurements and cross section and model evaluations. 2
The n TOF-ADS EC project
The n_TOF project [1] objective is the measurement, evaluation and dissemination of neutron cross sections relevant for the nuclear waste transmutation, ADS design and the development of the thorium cycle, including the topics of the first 4 points of the introduction. The project is organised as a shared cost action of the 5th framework program of the E.C, with wide participation of laboratories from many countries in Europe (Austria, France, Germany, Greece, Italy, Netherlands, Portugal, Spain and Sweden) plus CERN and the IRMM laboratory of the EC JRC. To perform the measurements a facility has been set-up at CERN [2] that will be initially exploited by an international collaboration, with the same partners of the EU contract plus several laboratories from Bulgaria, Poland, the Russian Federation and the USA. The IRMM facilities and other smaller accelerator will also be used to perform the experimental campaign. The measurements of this campaign will simultaneously provide useful information for the development of nuclear astrophysics, nuclear physics and neutron dosimetry. A summary of the n_TOF installation concept, design and present status is presented in another contribution to this conference [3] Several innovative detectors have and will be developed and installed in the experimental area. Parallel plate avalanche chambers, PPAC, observing the fission of standard isotopes, and advanced Si detectors plus one micromegas chamber observing (n,a) reactions in 10B and 6Li, will allow to obtain a precise absolute determination of the neutron flux intensity, energy distribution and beam profile [4], Additional BF3 detectors, working with the long counter principle, will allow the fast monitoring of these parameters. In a first phase, PPACs will also be used for the measurements and tagging of fissions in the samples. The capture reactions in the samples will be studied with newly designed C6D6 detectors [4]. In-beam spectroscopy with Ge detectors and
89
activation methods will be used for the measurements of (n,xn) reactions. All these advanced detectors will be handled by a fast, pileup resistant DAQ system, based on Flash ADC specifically designed for the experiment. In a second phase, scheduled for years 2003 and 2004, specific detectors will be installed to improve selection quality and maximum information recording of fission and capture events. The main component will be a 4rc calorimeter based on fast, neutron insensitive, inorganic scintillators [4]. This calorimeter is the key element for the capture cross section of the transuranic isotopes, where the separation of the capture events from fission and radioactive decay in the sample is a difficult task. 2.1
Detector commissioning
In parallel with the beam characterisation presented in [2], also contributed to this conference, specific measurements had been made to test the detector performance and proceed to a first calibration of the different measurement systems. ^.^^
- "y Flash" & relativistic particles _
Run: 1911 event: 1
250
200 fl> C £150 w
cu o a ioo
& so 0
500
1'000 1'500 2'000 2'500 3'000 3'500 4'000
Time [ns] Figure 5: Beginning of two typical histories, high (main picture) and low (insert) beam intensity, of one BICRON C«D« detector.
2.1.1
Capture measurements
During the first two years of the n_TOF experiment, capture measurements will be made by C6D6 detectors. Two versions of these detectors will be used and had been tested during the commissioning. First, four commercial BICRON detectors modified to reduce the neutron sensitivity and improve the response time. Second two special C6D6 detectors designed by FZK [4], with containers made of carbon fibre and that according to Monte Carlo simulations should further reduce the neutron sensitivity. Figure 5, shows two typical histories (high and low beam
90
intensity) of one of the CeD6 detectors. It can be observed that although the detectors are blind during 200ns by the flash of photons and relativistic particles, they recover fully within a few ps and are ready to acquire data in normal conditions at times corresponding to neutrons energies of several MeV. This energy range is well adapted for the measurements of capture and inelastic reactions for all isotopes. In addition the reduction of the background in the experimental area after the increase on the tunnel shielding at the end of year 2001, has reduced the width of the y flash and the associated background in the C6D6 detectors. r u n s j 730-1741 _C6D6_68
Figure 6: Raw neutron capture spectra from a typical run with a 1 mm thick the CiiDs detectors and a zoom around the 200 eV region.
Au sample, measured by
Samples of well known cross section have been measured with these detectors at n_TOF, notably Au, Ag and Fe [5,6]. Figure 2, shows the raw neutron capture spectra from a typical run with a 1 mm thick gold sample. These data have verified the time to energy calibration and the excellent energy resolution of n_TOF installation. However the demonstration of the full potential of n_TOF can not be done with a gold sample, since in this case the level density is so high at high energies that the resonances overlap each other. In order to extract the capture reaction rate from the spectra of raw C6D6 detector signals, each signal has to be weighted in such a way that the resulting photon detection efficiency is proportional to its energy. In this way the cascade detection efficiency becomes proportional to the known cascade energy and independent of the cascade path of each individual event. The n_TOF collaboration has demonstrated, that the best way to evaluate the weighting function is by a detailed simulation contrasted with experimental data. Figure 7 shows examples of the simulated gamma-ray C6D6 detectors response functions and of the deduced weighting functions in the presence of the Fe and Ag samples used in n_TOF. Figures 8 and 9 present the resulting weighted neutron capture spectra for the main resonance peaks of the Fe and Ag samples, respectively. The agreement to the fitted reaction rate profile obtained by SAMMY from the well known resonance
91 parameters and the Monte Carlo predictions of the neutron beam characteristics is excellent. GEANT4 - SIMULATED
SPECTRA
(5 MEvenis)
Fe 3 . 5 n m x Ag
Ziitrtrt
D . 2 m m X 20rrirn
'' . . 0
1
7
2
a
9
10
E0„„. [MeV]
Figure 7: Example of the simulated gamma-ray dDe detectors response functions and of the deduced weighting functions.
1.25 Energy (keV) Figure 8: Weighted capture spectra and SAMMY fit for the 1.15 keV Fe resonance peak.
The energy resolution of the n_TOF installation has not been determined yet, however the measurements of the 1.15 keV Fe capture resonance, Figure 8, show a 4 eV FWHM width. The continuous line shows the resonance fit with the SAMMY program [7] including the effects of the Doppler broadening, the C6D6 weighting functions and the Monte Carlo simulated n TOF resolution function. From this
92
fitting we estimate that less than 1 eV (0.1% FWHM) can be attributed to the n TOF resolution function. | A g 5.2eV 0 . 2 m m x 2 0 m m
J
I
2
I
I
I
I
I
I
4
I
I
I
,
6
I
I
8
I
l__l
I
10 Energy [eV]
Figure 9: Weight capture spectra and SAMMY fit for the Ag main resonance peak.
The data already presented in Figures 8 and 6, corresponding to capture in Fe and Au respectively show the presence of a non negligible background between the capture resonances in the data from the first data taking periods (before summer 2001). The background level was not very different from what is observed in the other existing neutron TOF facilities, but we expected to reach a much better performance because of the unique combination at n T O F of a very long flight path (185m) and a very low duty cycle (0.06 - 0.8 Hz). (n,y) in —'*—t—
i97
A u (o.i mm thick)
nTOF measured in J i j n e 2 0 0 1 ! n T OF n " a ^ u r c d in O c l m ' H ^ ( P 1
'
i
rL;
I
*•
i**W«+
Si , i , . . 1 . , 120
140
J\T
1 , , , 1 , 160
160
200
220
Er,feV)
Figure 10: Gold counting rates before and after the October 2001 shielding of the nTOF tunnel.
93
A large number of measurements were devoted to clarify the source of this background. The conclusion was that, for the background observed lus after the time of flight of photons, the most probable source are high energy muons and neutrons generated in the surroundings of the lead spallation target and travelling through the n_TOF tunnel to the experimental area. These high energy particles can also generate, after moderation and capture in the concrete walls of the experimental area, the background observed in the capture measurements at much longer times (corresponding to lower energies). An increase of the shielding near the spallation target was proposed to reduce substantially the observed backgrounds in the experimental area. After implementing a large additional shielding (3m of iron) at the nTOF tunnel in October 2001, the background in the experimental area was drastically reduced, the measurements show a reduction of more than a factor 5 close to the irradiated sample, Figure 10, and more than a factor 10 far from the sample, showing that the remaining background is mainly sample dominated. At the end of year 2003 and during 2004, capture cross sections will be performed with a high performance, 4n total absorption y calorimeter, inspired on the design of existing BaF2 FZK calorimeter, Figure 11. This detector will be mainly devoted to capture measurements of fissionable isotopes (actinides) and highly radioactive samples. The main advantages with respect to C6D6 detectors will be: the possibility to separate capture from fission events; the discrimination between capture events and other sources of y's (radioactive decay, background, etc..) by measuring the total cascade energy, a better control of systematics (no need of weighting function or similar corrections and in addition the tails of the total energy absorption allow to estimate the efficiency and contamination); and a higher absolute event efficiency where we expect to improve by a factor 5 with respect to the present C6D6 detectors.
Figure 11: General concept, single crystal BaF2 crystal and 4m FZK calorimeter pictures.
94
2.1.2
Fission measurements
The performance of the fission detectors and the capabilities of the nJTOF facility for measuring fission cross sections was also tested with samples of standard cross sections, namely 235U, 238U and 209Bi [6]. The counting rate was analysed, contrary to what will be done normally, to extract the cross section from a flux (a rough flux estimation from Monte Carlo simulations). Figure 12 shows the comparison of the obtained apparent cross sections with the databases. This figure shows that with the same setup it is possible to measure the fission cross section for energies ranging from leV to several hundred MeV (nearly 9 orders of magnitude). It also shows that the data reproduce correctly the fission thresholds and the characteristic energies for opening of the (n,nf) and (n,2nf) reactions in the U isotopes. In general the absolute value of the cross section (actually of the effective fluence) was correctly reproduced even by the rough Monte Carlo, within a factor 2. A more detailed simulation and analysis in progress, should substantially improve this agreement. .-235 — 235
U ENDF-B.6f°'o *>»w INDC(NDS)368«<"°2SOM»I/) U(n,f) n_TOF Exp. Data * 0 9 Biri_TOF Exp. Data • 2 M BI INDC(NDS)368
10
Figure 12: Comparisons between experimental data for sections
3
5
U,
238
10 10 10 Neutron Energy (eV)
U and
209
Bi and the evaluated cross
Program of neutron cross section measurements at n_TOF
A large number of measurements will be made at the n_TOF facility, after the beam characterisation and the detector commissioning is finished. These measurements will include the following measurements related to the Thorium fuel cycle, to transmutation and to ADS design: • 232 Capture cross sections of Th-cycle isotopes: Including measurements of Th, 23l 233 234 236 Pa, U, U and U in the range from leV to ~1 MeV. These isotopes will be measured first with C6D6 during 2002 and 2003, although some measurements will n
95 be repeated with the 4rc calorimeter by the end of 2003 and 2004. It should be noted that U (70y), that is a very interesting isotope for the thorium cycle, has a very high y specific activity, and depending on safety considerations and sample and proton availability it could be measured with the 47t calorimeter by the end of 2004. On the other hand, m Pa (1.3Id) and 233Pa (27d) although also interesting are too short lived to be directly measured at nTOF and finally 230Th (7.5x104y) measurements are not considered at present. Capture cross sections of transuranic isotopes: Pu, 42Pu, 2 'Am, 2 3Am and 245 Cm will be measured at nTOF mainly in 2004 with the 4rc calorimeter and 237Np will be measured by the nTOF collaboration at IRMM at Geel. The aimed energy range of measurements will be from leV to approx. 1 MeV. 238Pu (87.7y) has a very high specific activity, and depending on safety considerations and sample and proton availability, it could be measured with the 4% calorimeter by the end of 2004. Although several other isotopes are interesting targets, limitations in the sample availability and intrinsic radioactivity as well as proton availability, have reduced the present list of considered samples. In particular, 242Cm (163d), 243Cm (29y) and 244 Cm (18y) are too short lived and have high specific activities, and also no measurements are considered, at present, for241Pu(14.4y), 244Pu (8xl07y), 242m Am (141y), 246Cm (4760y), 247Cm (1.6xl07y) and 248Cm (3.5xl0sy). Capture cross sections of Long Lived Fission Fragments: 151Sm will be measured by nTOF at CERN (nTOF-03) with C6D6 detectors during 2002. l51Sm is a medium lived fission fragment (90y) with additional interest for astrophysics. On the other hand, 99Tc (2.1xl05y) and l29I (1.6xl07y) are the LLFF with highest impact on the long term radiotoxicity inventory and the two with highest possibilities of reaching technical feasibility of transmutation. They are/will be measured by nTOF at IRMM at Geel. 79Se (6.5x104y) could be measured depending on sample and proton availability. Finally, Sn (10 y), Cs (2.3x10 y), 93 Zr (1.5xl06y), 107Pd (6.5xl06y) are difficult samples to obtain and no measurements are considered at present. In addition, their transmutation requires, in most cases, difficult isotopic separation procedures. Capture cross sections of Coolants and Structural materials: Lead and Pb/Bi alloys have an important role in ADS, both as the target of the spallation source and as coolant of the nuclear core. Po produced by Pb and Bi activation is a source of concern from the point of view of radioactive wastes and radiation protection in ADS operation. Furthermore, Pb and Bi isotopes will contribute to the neutron balance and spectrum definition by capture, elastic and inelastic scattering and by the (n,xn) reactions. As a consequence, 204.206.207.208pb, 209Bi will be measured by nTOF at CERN (PI42) with C6D6 detectors during 2002 and 2003. Fission cross sections of Th-cycle and transuranic isotopes: The main isotopes considered are 237Np, 239Pu, 241Am, 243Am, 244Cm, 245Cm, 232Th, 23,Pa, 233U, 234U and 236 U (plus 235U and 238U - reference standard isotopes). The main objective will be to cover the energy range from leV to 20 MeV, but the higher energy limit will be extended as much as allowed by statistics. Fission measurements are very time
96
consuming and, although the present PPACs already foresee the possibility to use 11 detectors, two additional actions are being studied to improve proton utilization for fission measurements: the use of additional multi target fission detectors; and the construction of a second experimental area, probably before the start of the 2003 measurement campaign, that would enable to perform simultaneously fission and capture measurements. Total cross sections: Performed by transmission, most probably in the IRMM facilities. The presently proposed isotopes are Np, I, Pu and Pu. (n,xn) cross sections: Performed in two ways, by TOF at CERN and by activation methods in several facilities at Europe providing monoenergetic neutrons. Adding together both types of installations, measurements are proposed for Np, 232 Th, 23lPa, 239Pu,24lPu, 24,Am, 243Am, 233U, and 207Pb. In addition, there will be astro-nuclear and basic nuclear physics cross section measurements. The facility is expected to operate at CERN for a much longer period of time after the EC contract, and the previous list of measurements is expected to grow later on. References 1.
The Contract FIKW-CT-2000-00107 for the "ADS nuclear data (N-TOF-ADS)" EC project. Signed on October 2000. 2. .S. Andriamonje et al., "Neutron TOF Facility (PS213) Technical design Report". CERN/INTC/2000-004. February 2000. 3. A. Mengoni on behalf of the n-TOF collaboration. Contribution to these proceedings 4. The n_TOF Collaboration, "The nTOF Technical Report", CERN/INTC 2000018, November 2000. 5. E. Gonzalez et al., "Preliminary report of the first nTOF detectors commissioning period', CERN n T O F Note. April, 2001. 6. The n T O F Collaboration, "Status Report [July 2001J" (of the n T O F project). CERN/INTC 2001-021, August 2001. 7. N.M. Larson, "SAMMY: Multilevel R-Matrix fits to neutron data using Bayes equations", ORNL/TM-9179, Oak Ridge NL, 2000.
ACTIVITIES OF THE DOE NUCLEAR CRITICALITY SAFETY PROGRAM (NCSP) AT THE OAK RIDGE ELECTRON LINEAR ACCELERATOR (ORELA) TIMOTHY E. VALENTINE, LUIZ C. LEAL, AND KLAUS H. GUBER Oak Ridge National Laboratory, P. O. Box 2008, Oak Ridge, 77V 37931 USA E-mail: [email protected] The Department of Energy established the Nuclear Criticality Safety Program (NCSP) in response to the Recommendation 97-2 by the Defense Nuclear Facilities Safety Board. The NCSP consists of seven elements of which nuclear data measurements and evaluations is a key component. The intent of the nuclear data activities is to provide high resolution nuclear data measurements that are evaluated, validated, and formatted for use by the nuclear criticality safety community to provide improved and reliable calculations for nuclear criticality safety evaluations. High resolution capture, fission, and transmission measurements are performed at the Oak Ridge Electron Linear Accelerator (ORELA) to address the needs of the criticality safety community and to address known deficiencies in nuclear data evaluations. The activities at ORELA include measurements on both light and heavy nuclei and have been used to identify improvements in measurement techniques that greatly improve the measurement of small capture cross sections. The measurement activities at ORELA provide precise and reliable high-resolution nuclear data for the nuclear criticality safety community.
1
Introduction
The Defense Nuclear Facilities Safety Board (DNFSB) Recommendation 97-2 motivated the U.S. Department of Energy (DOE) to establish the Nuclear Criticality Safety Program (NCSP). The NCSP consists of the following seven tasks: critical and subcritical experiments, criticality benchmarks, training, analytical methods, data preservation, applicable ranges of bounding curves and data, and nuclear data measurements and evaluations. This comprehensive program was developed to maintain the fundamental infrastructure to retain the capabilities and expertise in nuclear criticality safety. The nuclear data task for the NCSP is focused on measurement and evaluations of cross sections that are needed for the nuclear criticality safety program. Differential cross section measurements are performed at the Oak Ridge Electron Linear Accelerator (ORELA)[l] at Oak Ridge National Laboratory (ORNL) while the analysis and evaluation of these measurements are performed using the generalized least-squares fitting code SAMMY[2] in the resolved, unresolved, and high energy regions. The evaluation includes recent ORELA measurements and past measurements to generate a complete evaluation for the nuclides of interest for inclusion into the Evaluated Nuclear Data Files (ENDF/B). The evaluations are
97
98
processed using the NJOY[3] and AMPX[4] codes to produce code dependent nuclear data files for validating the nuclear data evaluations. The evaluations are submitted to the Cross Section Evaluation Working Group (CSEWG) that is administered by Brookhaven National Laboratory (BNL) for validation and release to the user community. In this paper, we discuss the motivation for performing cross section measurements at the ORELA. We provide a review of the recent measurement activities at ORELA and discuss the need for future measurements at ORELA and other facilities. Finally, we summarize a few general observations resulting from these activities. 2
The Need for Nuclear Data Measurements
The objective of nuclear criticality safety is to ensure that fissile material is handled in such a way that it remains subcritical under both normal and credible abnormal conditions to protect workers, the public, and the environment. Credible abnormal conditions include but are not limited to the addition of moderating materials into fissile packages and changes in reflection, moderation, and absorption in fissile packages. Criticality safety evaluations involve the use of neutron transport codes to calculate integral quantities that are directly dependent on the adequacy of the nuclear data. The key parameter used to characterize the safety of fissile packages is the effective neutron multiplication factor, keff. The effective neutron multiplication factor relates the production of neutrons in an assembly to the losses in the assembly and is defined as:
\\v{E)Lf (r, E)(r, E)dVdE keff ~
^OOPNL
- ^NL
\\La(r,E)(r,E)dVdE
Xi(r,E) = N{r)ai(r,E). In this expression, PNL is the non-leakage probability, v(E) is the number of neutrons from fission, 2f(r,E) is the macroscopic neutron fission cross section, Sa(r,E) is the macroscopic neutron absorption cross section, <j>(r,E) is the neutron flux, a(r,E) is the microscopic cross section, N(r) is the atom density of the material, r is the spatial variable, and E is the energy variable. As can be seen from the above expression, the computation of keff is directly dependent on the nuclear data. The accuracy of the computation of keff is in most part directly dependent on the accuracy of the nuclear data and to a much lesser extent on the limitations of the transport models. Historically, much of the nuclear data measurement activities were driven by the fast and thermal reactor programs and by the fusion programs. The thermal reactor
99
programs were interested in nuclear data for neutrons with energies of a few electron volts whereas the fast reactor and fusion programs were mainly concerned with neutrons whose energies were in the MeV region. Much of the data in the intermediate energy region were not of a concern for the reactor programs; therefore, high-resolution data in this region was not of a concern. In recent times, the data in the intermediate energy region (1 eV to 1 MeV) has become much more important for nuclear criticality safety applications. Criticality safety analyses of storage and transport conditions for spent nuclear fuels along with analyses of the decommissioning and decontamination of legacy nuclear facilities involve conditions in which the neutron flux in these systems have a significant component in the intermediate energy region. Therefore, the adequacy of the nuclear data in the intermediate energy region is of greater concern. The detailed structure of the resolved resonance region is needed to adequately model the resonance structure including windows in the cross sections, is needed to compute resonance self shielding for generating multi-group cross section data files, is needed to produce temperature dependent nuclear data files, and is needed to calculate the unresolved resonance cross sections. 3
Measurement Activities at ORELA
The ORELA consists of a 180 MeV electron linear accelerator, neutron producing targets, buried and evacuated flight tubes up to 200 m long leading to underground detector locations, sophisticated detectors, and data acquisition systems. Simultaneous measurements are possible at the 18 detector stations on 10 separate flight paths. The ORELA is the only facility available in the United States that can be used to provide high-resolution nuclear data from thermal to intermediate energies (10 5 to 106 eV). Consequently, there are only two other facilities in the world that have comparable capabilities. 3.1
Transmission Measurements
Transmission measurements are commonly preformed at the 80-m flight path using a 6 Li-glass scintillation detector. The samples are placed approximately 9 meters from the neutron source while the detectors are located 80 meters from the neutron source. The transmission measurements are typically performed at two different pulse repetition rates (525 Hz and 130 Hz). Measurements at 525 Hz are performed to acquire the most data in the intermediate energy region in the shortest amount of time while measurements at 130 Hz are performed to correct for the overlap of pulses that result from slow neutron from a previous pulse appearing as fast neutrons in a subsequent pulse. The measurements are typically performed with four different samples that include the sample in a sample holder, an empty sample holder, a polyethylene sample, and a "black" sample. The purpose of the polyethylene sample
100
is to measure the time-dependent background due to neutron capture in the target moderator while the "black" sample is used to provide a correction to ensure that large resonances have zero transmission. Various filters can be inserted before the sample to tailor the neutron spectrum and reduce the gamma flash that results from the electrons hitting the tantalum target. Transmission measurements have been performed at ORELA for several nuclides of interest for nuclear criticality safety. The measurements include 233U, aluminum, natural chlorine, and natural potassium. The 233U sample was cryogenically cooled using liquid helium to reduce the effects of Doppler broadening. A comparison of the recent 233U transmission measurements performed at ORELA[5] and those performed by Pattenden et al[6] is provided in Fig. 1. As can be seen from this figure, the ORELA data is much more resolved than the older measurements that were performed at a reactor using a chopper. Hence, the computation of the resonance self-shielding and the neutron slowing down will be much more accurate.
600
ORELA 1998 Pattenden et al. 1963
500
60
70
100
Energy (eV) Figure 1. Comparison of ORELA and Pattenden et al total cross section measurements for
U.
101
3.2
Fission Cross Section Measurements
Fission cross section measurements were performed for 233U for nuclear criticality safety[7]. The 233U fission cross section measurements were performed with a multiplate fission chamber at the 80-m flight station. The fission chamber contained 2.11 grams of Uranium that was enriched to 99.997 wt% 233U. The chamber was comprised of 21, 0.127-mm-thick aluminum plates that were coated on each side, except the end plates, with U 3 0 8 that was 76 mm in diameter. A sketch of the fission chamber is provided in Fig. 2.
J " ' n - 00 PfLLING TU8E
t, NYICN SNJ05 AN!? NUTS
0
I
i
3
SCMjE IN INCHES
Figure 2. Sketch of the multi-plate 233
U fission chamber.
The U fission cross section measurements were performed at repetition rates of 400 Hz and 78 Hz with a neutron pulse width of 8 nanoseconds. The 400 Hz measurements were performed to acquire data in the intermediate energy region while the 78 Hz measurements were performed to correct the 400 Hz measurements for the overlapping of low energy neutrons and to provide better counting statistics for the low energy fission cross section. In these measurements a thin 6Li-glass flux monitor was used to measure the neutron spectrum. A comparison of the ORELA fission cross section with the current ENDF/B-VI fission cross section for 233U is provided in Fig. 3. As can be seen from this figure, the ENDF/B-VI fission cross section is only resolved to 60 eV whereas the ORELA fission cross section is resolved to energies greater than 100 eV. The improved data above 60 eV will contribute significantly to the accurate calculation of resonance self shielding in the intermediate energy region.
102 233 233
in
U ORELA measurement 1997 UENDRB-VI evaluation
E n
50
60
70
80
90
100
Energy (eV) Figure 3. Comparison of the ORELA and ENDF/B-VI
3.3
U fission cross section.
Capture Cross Section Measurements
Neutron capture cross section measurements are performed at the 40-m flight path station. The capture measurement detector configuration has been greatly improved compared to the original system used to perform neutron capture cross section measurements. The former capture system contained too much structural material in the vicinity of the detectors. This large amount of structure produced an energy dependent gamma ray background due to neutrons that scattered from the sample and were subsequently captured in the structural materials. Attempts were made to correct for the effects of scattered neutrons, but these corrections have been shown to be inadequate[8]. Furthermore, the gamma ray detectors were C6F6 detectors that were also sensitive to neutron detection whereas the C6D6 detectors that are now used to perform the measurements are much less sensitive to neutrons. The capture cross section measurements are performed with a neutron pulse width of 8 nanoseconds and with repetition rates of 525 Hz and 130Hz. Measurements are performed at a 525 Hz repetition rate to reduce the measurement times required to achieve the desired counting precision. The 525 Hz measurements are performed with the sample of interest, with an empty sample holder, and with the sample of interest replaced with a carbon sample. The measurement with the empty sample holder is used to determine a background correction due to the ambient gamma rays in the environment while the carbon sample is used to determine the correction required to account for gamma rays produced from neutrons that scatter off of the sample and are subsequently absorbed in the environment around the
103 capture system. A measurement is performed with a repetition rate of 130 Hz with the target of interest replaced with a thin gold foil target. The 4.9 eV resonance in gold is used as a reference to provide a normalization factor that is applied to the data for the sample of interest. The lower repetition rate is required in order to measure the 4.9 eV resonance in gold. Capture cross section measurements have been performed for several samples of interest for nuclear criticality safety including aluminum, natural chlorine, natural silicon, fluorine, natural potassium, and potassium enriched in 4IK. As a example, a plot of the chlorine capture cross section is provided in Fig. 4[9]. The plot of the data in Fig. 4 demonstrates how well the data can be fitted using SAMMY.
fit
.-.
st.00 0.4 0.. 3 20..1 0.
-i—<—i—>—i—>—i—'—r~'—r—'—i—'—i—'—i—'—r—••—r 105 1.1.0 I I S 120 125 130 13S 140 145 150 15S v '
I
T
T
i
-
D
'
r
>fc J —
35
40
•*
30
25
I
t
1
t'
16
*
1
SO
45
2, i 1 0. 0. 1 100,14.0 10 1 0.1 0,01
1
14.5
i
15.0 i
0 38
15,5 i
— p
16.0
0 40
i
165 |
1.7,0 i
0 42
|
0 44
17.5
i
18.0 |
0.46
Energy (keV) Figure 4. Plot of measured capture cross section for natural chlorine.
A comparison of the ORELA and the latest ENDF/B-VI capture cross sections for silicon isotopes is provided in Table 1 [10]. As can be seen from this table, the latest ORELA neutron capture cross section data is in general smaller than the ENDF/BVI data. In general the recent ORELA capture cross sections data for many of the light nuclei are smaller than the currently available data the ENDF/B-VI data. This difference is attributed to the much higher sensitivity of the older capture apparatus to scattered neutrons[8]. The reduction in the capture cross sections will result in less neutron capture in systems that contain silicon, chlorine, aluminum, fluorine, and potassium. If the systems contain large amounts of any of these materials, then
104
the computed keff value will be larger using the latest ORELA data as compared to the results obtained using the current ENDF/B-VI data. The recent ORELA data may impact criticality safety analyses of storage and transport of fissile materials especially for systems that contain a significant component of the neutron flux in the intermediate energy region. Table 1. Comparison of ORELA and ENDF/B-VI capture cross sections for silicon. Energy (keV) 2.5e-5 -1 1-5 5-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-150 150-200 200-250 250-300 300-350 350-400 400-450 450 - 500 500 - 550 550-600 600 - 650 650 - 700 3.4
28
29 Si Si (mb) (mb) ORELA ENDF/B ORELA ENDF/B 9.95 9.81 7.05 6.96 0.48 0.53 0.38 0.38 0.25 0.33 0.23 0.23 0.17 0.24 22.1 30.51 0.10 0.21 0.15 0.15 0.61 0.84 30.8 32.51 0.10 0.32 0.44 0.41 1.37 12.05 0.10 0.10 7.94 16.90 0.08 0.08 0.29 0.72 0.07 0.07 1.79 5.69 0.07 0.07 0.11 0.20 0.06 0.06 0.31 0.37 0.07 0.07 1.31 1.69 2.09 1.34 0.29 0.38 0.05 0.05 0.40 0.76 0.06 0.05 0.02 0.33 0.33 0.23 0.49 0.74 1.33 0.06 0.01 0.04 0.15 0.16 0.01 0.04 0.31 0.06 0.29 0.92 0.19 0.20 0.66 2.36 0.80 1.19 0.69 1.64 0.93 1.13 0.02 0.06 0.38 0.55
30
Si (mb) ORELA ENDF/B 6.28 6.58 42.5 43.9 0.55 1.06 2.85 0.15 0.11 0.11 0.10 0.10 0.09 0.09 0.09 0.08 0.08 0.08 0.09 0.08 0.09 0.08 0.09 0.08 0.17 0.13 3.64 2.49 0.97 0.75 0.05 0.05 0.23 0.23 0.02 0.02 0.13 0.14 0.01 0.01 0.01 0.01 0.01 0.01 0.10 0.22 0.01 0.02
A dditional Measurement Needs
Both transmission and capture measurements are needed for many nuclides to address additional criticality safety concerns. In particular, measurements need to be performed for fissile material processing operations that include lithium, potassium, and sodium carbonates. Measurements need to be performed for nitrogen and
105 phosphorous to address data needs associated with process residues and deposits that are encountered in various nuclear facilities. Measurements are also needed for constituents of structural materials to reduce conservatisms associated with the criticality safety analyses of storage and shipment of spent nuclear fuels. All of these data needs can be addressed by performing measurements at ORELA. Fission products in spent fuel packages absorb neutrons and thus reduce the reactivity of these packages. However, in most criticality safety analyses the impact of the fission products is not taken into account when establishing fuel storage and transport limits. In the future, measurements of radioactive fission products may be necessary to reduce conservatism associated with storage and transport of spent nuclear fuel. Due to the high brightness of the ORELA neutron producing target, ORELA would be an excellent facility for transmission measurements on radioactive samples whereas the SNS would be the best facility for performing neutron capture measurements because the high neutron flux is the most important parameter for such measurements. 4
Summary
The DOE NCSP was established to maintain the basic infrastructure to support nuclear criticality safety operations in the DOE complex. Nuclear data forms the basis for all neutron transport calculations for nuclear criticality safety; therefore, one element of the NCSP consists of measurement and evaluation of nuclear data. The recent challenges faced by the nuclear criticality safety analysts involve configurations of fissile materials in which the nuclear data in the intermediate energy region are of greater importance. High-resolution neutron cross section measurements are performed at ORELA to address the needs and provide improved data in the intermediate energy region. The latest transmission, fission, and capture cross section measurements performed at ORELA have provided significantly improved nuclear data for criticality safety. The greatest differences have occurred in the capture cross sections for light nuclei. In several of the measurements with light nuclei, the capture cross sections over some energy regions have been shown to be between 2 to 10 times smaller than the currently available data in the evaluated nuclear data files. Additional measurements are still needed to address conservatisms associated with cleanup of various nuclear facilities and with shipment of spent nuclear fuels. Many of the current needs can be addressed at ORELA although capture measurements on radioactive samples would be best performed at the Spallation Neutron Source.
106
5
Acknowledgements
ORNL is managed by UT-Battelle, LLC for the U.S. Department of Energy under contract number DE-AC05-00OR22725. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
R. W. Pelle et al., Technical Report No. ORNL/TM-8225, Oak Ridge National Laboratory. N. M. Larson, ORNL/TM-9179/R5, Oak Ridge National Laboratory (2000). R. E. MacFarlane and D. W. Muir, LA-12740-M, Los Alamos National Laboratory (1994). N. M. Greene et al., ORNL/TM-3706, Oak Ridge National Laboratory (1976). K. H. Guber et al., Nucl. Sci. Eng., 139, 111 (2001). N. J. Pattenden and J. A. Harvey, Nucl. Sci. Eng., 17, 404 (1963). K. H. Guber et al., Nucl. Sci. Eng., 135, 141 (2000). P. E. Koehler et al, Phys. Rev. C 62, 055803 (2000). K. H. Guber et al, Phys. Rev. C 65, 058801 (2002). K. H. Guber et al, "Neutron Cross Section Measurements for Light Elements at ORELA and their Application in Nuclear Criticality," Proc. Nuclear Data Conf., Tsukuba, Japan, October 2001.
P A R A M E T E R S FOR N U C L E A R R E A C T I O N CALCULATIONS - N E E D S FOR I M P R O V E M E N T S M. HERMAN International Atomic Energy Agency, Vienna,
Austria
The status and contents of the Reference Input Parameter Library (RIPL) are summarized. This input library has been developed for theoretical calculation of nuclear reactions, and suggests experimental activities at the SNS that could bring improvements to the data and consequently increase the accuracy of model calculations.
1
Introduction
Increased use of nuclear reaction theory for predicting cross sections, spectra, and angular distributions, as required for a large variety of application, is an important trend in the evaluation of neutron and charged-particle nuclear data. The model codes offer important advantages such as ensuring internal consistency of the data by preserving the energy balance and the coherence of the partial cross sections with the total or the reaction cross sections. In addition, theoretical calculations represent the only approach that can fill gaps in the experimental results and predict data for unstable nuclei. Nuclear astrophysics and the design of Accelerator Driven Systems are typical applications that depend strongly on theoretical calculations. The nuclear reaction theory is believed to be in a position to meet many of the requirements for practical applications. The major sources of uncertainty are, the input parameters needed to perform theoretical calculations. The IAEA has addressed these needs through a Coordinated Research Project on the Reference Input Parameter Library (RIPL), which involves the difficult task of collecting, evaluating and recommending the vast amounts of various nuclear parameters. The first phase of the project was completed in 1999, with the production of a Starter File and related documentation (TECDOC10341). A second phase of the project was initiated in 1999 to test the RIPL-1 database and produce interfaces between RIPL and commonly used nuclear reaction codes. Substantial improvements and extensions to the original database have been made. The RIPL-2 is expected to be released in July 2002. The contents of the RIPL-2 library are outlined below, with possible improvements that could be made to the current database through new measurements at the SNS.
107
108
2 2.1
C o n t e n t s of RIPL-2 Segment 1: MASSES
The mass segment contains basic ground state properties of nuclei, along with two theoretical predictions of masses and deformations. The first one is based on the Hartree-Bock-Bogolubov (HFB) theory and was fitted to all 1888 measured masses of nuclei with N and Z > 8. The second file contains predictions obtained within the Finite Range Droplet Model (FRDM) 2 . The atomic mass excesses and nuclear ground-state deformations are tabulated for 8979 nuclei ranging from 1 6 0 to A=339. The most recent evaluated experimental masses by Audi and Wapstra 3 are included. A third possibility is provided by a subroutine implementing the Duflo-Zuker formula4 which reproduces the 1950 experimental masses above 4 He with an rms error of 574 keV. 2.2
Segment 2: LEVELS
This segment contains 110 files (one for each element) with all known level schemes available from ENSDF in 1998. These files are arranged and preprocessed into an easy-to-read format for nuclear reaction codes. During preprocessing all missing spins were inferred uniquely for each level from spin distributions extracted from the existing data. Electromagnetic and 7-ray decay probabilities were estimated. Missing internal conversion coefficients (ICC) were calculated using the inferred or existing spin information. Particle decay modes are also given whenever measured. This segment also contains the results of constant temperature fit of nuclear level schemes intended for use in the level density estimation. 2.3
Segment 3:
RESONANCES
The average resonance parameters recommended for RIPL-2 were prepared on the basis of the evaluations performed by the Obninsk group, taking into account the analysis of discrepancies between similar evaluations of other groups. Good agreement was found for T 7 among the three RIPL-1 files (Obninsk, Mughabghab and Beijing)1. The revised average resonance parameters were obtained for 20 additional nuclei for which the data on resolved resonance parameters are available in the Sukhoruchkin compilation 5 , bringing total number of D0(,s in RIPL-2 to 301. New evaluations of the average parameters for p-wave neutron resonances, prepared by the Obninsk group, have been included in the updated version of the RIPL-2 file. Careful attention was paid to the estimation of uncertainties for the rec-
109
ommended parameters, based on experienced guesswork of systematic errors beside statistical uncertainties. 2.4
Segment 4:
OPTICAL
The optical model parameter (OMP) segment is provided in two forms: full library (archival form) and shorter library with all single-energy potentials removed (user file). Currently, the user (archival) file contains 258 (258) potentials for incident neutrons, 98 (143) potentials for incident protons, 8 (11) for deuterons, 1 (26) for tritons, 3 (53) for 3 He particles, and 10 (10) for incident a-particles. Of the neutron potentials, 229 are spherical potentials, 28 are coupled-channels potentials, and 1 is a vibrational model. There are 6 coupled-channels and 92 spherical potentials for incident protons in the user library, and a total of 5 dispersive optical potentials for incident neutrons. The new global potential for neutrons and protons from Koning and Delaroche6 was incorporated, as were new dispersive potentials from Capote. Where there are not enough experimental data to define phenomenological OM parameters, one has to resort either to global parameterizations or to new microscopic approaches. The semi-microscopic model developed at Bruyeres is now part of the OM segment, incorporating a revised version of the MOM code which relies on the Jeukenne, Lejeune, and Mahaux nuclear matter approach. A compilation of 1708 deformation parameters (/32 and /J3) for collective levels has been retrieved from the JENDL-3.2 evaluations, ENSDF and literature to be used in direct reaction calculations. 2.5
Segment 5: LEVEL
DENSITIES
The total level density sub-segment contains a revised version of the Back Shifted Fermi Gas (BSFG) model parameters prepared by the Obninsk group, which are consistent with both the recommended RIPL-2 neutron resonance parameters and the evaluated parameters of the recommended low-lying levels. The new BSFG systematics developed by the Brussels group is also consistent with the recommended RIPL-2 neutron resonance parameters, and will be included in the RIPL-2 TECDOC. The Gilbert-Cameron (GC) and Generalized Super-fluid Model (GSM) parameters were revised by the Obninsk group in accordance with changes in the RIPL-2 resonance segment. The microscopic HF-BCS calculations of the nuclear level densities are based on the realistic microscopic single-particle level scheme7 and were supplied by Goriely. Also, the single-particle schemes used in the HF-BCS calculations were provided by the Brussels group. In addition, the FRDM single-particle schemes are included as corresponding to the accepted FRDM mass table.
110
A critical review was undertaken of the methods for calculating partial level densities to be used in preequilibrium model calculations. A code for combinatorial calculation of particle-hole state densities, based on a convolution of shell-model single particle-states with BCS pairing, is included in RIPL-2 along with the corresponding tools for retrieving singleparticle levels from Segment 1. The most useful analytical approaches in the frame of the equidistant single-particle model are implemented in the revised AVRIGEANU code 8 . 2.6
Segment 6: GAMMA
This segment contains parameters that quantify Giant Resonances, experimental 7-strength functions and methods for calculating 7-emission in statistical model codes. The experimental Giant Dipole Resonance parameters were provided by the Chinese group, as represented by Lorentzian fits to the total photo-neutron cross sections for 102 nuclides ranging from 5 1 V to 2 3 9 Pu as compiled by Dietrich and Berman 9 . Additional data were estimated by Liu Jianfeng and Su Zongdi in 1995 10 . New compilations of calculated GDR widths and energies for about 6000 nuclei with 14
Segment 7: FISSION
Fission is a new RIPL-2 segment, which retains the RIPL-1 recommendation and, in addition, includes global prescription for barriers and nuclear level
111
densities at saddle points. Fission barrier parameters for the trans-thorium nuclei were recommended by Maslov1 and for the preactinides by Smirenkin 16 . For nuclei with Z<80 the liquid drop barriers described by Sierk's code 17 are recommended with the addition of the ground-state shell corrections estimated by the MoellerNix (Segment 1) or the Mayer-Swiatecki (Segment 5) mass formulae. Another option is to predict the fission barriers and saddle point deformations obtained within the Extended Thomas-Fermi plus Strutinsky Integral (ETFSI) method of Goriely. Experimental primary barriers can be reproduced within plus or minus 1.5 MeV. The present ETFSI compilation includes 2301 nuclei with 78
Testing
Tests have been performed on the optical, resonance and levels segments. A number of misprints and erroneous coding have been detected and corrected. Several RIPL participants tested the preliminary version of the levels database by using the data in calculations. A new simple test was worked out for checking nuclear temperature (T) derived from the analysis of cumulative plots of discrete levels, yielding temperature values which are remarkably similar to the T(A) function obtained in the global fitting procedure. Herman has extensively tested N m a x values for nearly 500 nuclei using the GilbertCameron procedure and level densities specific to the EMPIRE code. Global testing of the RIPL-2 database has been performed in three separate exercises. Large numbers of nuclear reaction cross sections were calculated by means of the nuclear model codes EMPIRE-II, UNF and TALYS. Herman performed calculations for the most important neutron-induced reactions on 22 targets from 4 0 Ca up to 2 0 8 Pb in the energy range from 1 keV up to 20 MeV. Comparison with experimental data shows reasonable overall agreement for most of the calculations. There is a clear indication that calculations using the new RIPL-2 files fit experimental data better than those with default EMPIRE-II parameters, which demonstrates the improvements brought about by RIPL-2. However, significant discrepancies among the results of the three sets of calculations were observed in a number of cases. These findings
112
illustrate the importance of the model parameters and prove the practical usefulness of the RIPL-2 library for basic research and applications. The second exercise was carried out by the Beijing group, using the recently developed UNF code to study 103 nuclei from the mass region 69-160 in the incident energy range from 0.1 to 20 MeV. All input parameters were taken from the RIPL database. Agreement with the experimental data was found to be very good for total and elastic cross sections (within 3%). For other main reaction channels, calculations reproduced the shape, but some parameter adjustments were necessary in order to fit the absolute cross sections. TALYS calculations were performed for various neutron-induced reactions on 5 isotopes from 5 2 Cr to 2 0 8 Pb. Default input parameters originated from RIPL-2. This exercise concentrated on the comparison of Ignatyuk-type and microscopic level densities and provided very reasonable agreement with experimental data for both formulations. 4
C o d e Interfaces
The work on interfaces between selected nuclear model codes and RIPL-2 segments has been facilitated by the standard RIPL-2 format. The two optical model codes (ECIS and SCAT2) and two statistical model codes (EMPIRE-II and UNF) use RIPL-2 library to a significant extent. Interface code preparing inputs for ECIS and SCAT2 have been written by Young and is available in the optical segment. The statistical model code UNF (PR China) makes use of RIPL optical potentials, masses, levels, level densities and GDR parameters. EMPIRE-II accesses RIPL-2 database directly and retrieves optical model parameters, discrete levels and microscopic level densities (HF-BCS). Built in systematics for GDR parameters and prescriptions for 7-strength functions follow RIPL-2 recommendations. EMPIRE-II library of masses and ground state deformations is numerically identical to the mass-frdm.dat in the mass segment of RIPL-2. TALYS uses a dedicated format for the input parameter library but numerical data are based on RIPL-2. 5
Conclusions
The RIPL-2 library is close to completion, with public release expected in July 2002. Users of reaction codes will benefit considerably from the generation of a complete and consistent set of starting parameters to give sensible results for cross sections and spectra. However, RIPL-2 should be further extended and continuously updated in order to retain the relevance and value of the library
113
to the users. At the recent co-ordination meeting in Vienna, December 2001 the CRP participants discussed possible improvements of the current project and formulated recommendations for further activities. These finding are summarized below: • RIPL-2 provides valid sets of parameters for spherical and near-spherical nuclei. On the other hand, data for the deformed nuclei are scarcer and less accurate. In particular there is a need for more Coupled-Channels potentials and 7-ray strength functions for the deformed nuclei. These can only be achieved if new experimental data become available. Such measurements (especially for neutron- and 7-strength functions) could be performed at high accuracy on the new SNS facility. • Special techniques should be applied for the determination of parameters for nuclei far from the stability line for which there are usually no experimental data available. These nuclei are important for ADS and astrophysics. The SNS facility, with its very high neutron flux, represents a unique source of experimental data on short-lived radioisotopes. • New experimental data from the recently initiated projects (HINDAS and N-TOF at CERN) should become available within a year or two, offering possibilities for testing RIPL-2 parameters. The same is true for the SNS facility at a somewhat longer time scale. • More attention should be dedicated to the use of microscopic models for producing parameters. Parameters related to the fission channel contained in RIPL-2 need more accurate analysis and improvement. • The problem of collective enhancement of level densities should be addressed in more detail in order to provide a reliable prescription for calculating level densities in deformed nuclei. The latter are often needed for ADS and new reactor concepts. 6
Participants
The following scientists contributed to the RIPL-2 library: T. Belgya (IISCCR, Budapest, Hungary), 0 . Bersillon (Bruyres-le-Chtel, France), R. Capote (NCEADNC, Havana , Cuba), T. Fukahori (JAERI, Tokai-mura, Japan), S. Goriely (Univ. of Brussels, Belgium), M. Herman (IAEA, Vienna, Austria), A. V. Ignatyuk (IPPE, Obninsk, Russia), S. Kailas (Bhabha, Trombay-Mumbai, India), A. Koning (Petten, Holland), P. Oblozinsk (BNL,
114
Brookhaven, USA), V. Plujko (Univ. of Kiev, Ukraine), P. G. Young (LANL, Los Alamos, USA), Ge Zhigang (CNDC, Beijing, China). References 1. Handbook for calculations of nuclear reaction data: reference input parameter library (International Atomic Energy Agency, Vienna, Austria, 1998, http://www-nds.iaea.or.at/ripl), No. IAEA-TECDOC-1034. 2. P. Moller, J. R. Nix, W. D. Myers, and W. J. Swiatecki, At. Data Nucl. Data Tables 59, 185 (1995). 3. G. Audi and A. H. Wapstra, Nucl. Phys. A595, 409 (1995). 4. J. Duflo and A. Zuker, Phys. Rev. C52, 23 (1995). 5. S. Sukhoruchkin, Z. N. Soroko, and V. V. Deriglazov, in Low Energy Neutrons Physics, Tables of Neutron Resonance Parameters, edited by H. Schopper (Springer-Verlag, Darmstadt, 2000), Vol. 16B. 6. A. Koning and J. Delaroche, to be published (2002). 7. S. Goriely, F. Tondeur, and J. Pearson, At. Data Nucl. Data Tables 77, 311 (2001). 8. M. Avrigeanu and V. Avrigeanu, Comput. Phys. Commun. 112, 191 (1998). 9. S. S. Dietrich and B. L. Berman, At. Data and Nucl. Data Tables 38, 199 (1988). 10. Liu Jianfeng and Su Zongdi, Chinese J. Nucl. Phys. 17, 336 (1995). 11. P. Van Isacker et al., Phys. Rev. C45, R13 (1992). 12. S. Goriely, Phys. Lett. B436, 10 (1998). 13. F.-K. Thielemann and M. Arnould, in Conf. on Nuclear Data for Science and Technology, Antwerp, 6-10 September, 1982, edited by K. Bockhoff (Reidel, Dordrecht, The Netherlands, 1983), p. 762. 14. S. Goriely and E. Khan, Nucl. Phys. A submitted for publication (2002). 15. E. Khan et al., Nucl. Phys. A694, 103 (2001). 16. G. Smirenkin, Report INDC(CCP)-359, IAEA, Vienna (unpublished). 17. A. J. Sierk, BARMOM, no.967 ed., National Energy Software Center (Argonne National Laboratory, IL60439). 18. A. Mamdouh, J. Pearson, M. Rayet, and F. Tondeur, Nucl. Phys. A679, 337 (2001).
ALUMINUM DATA MEASUREMENTS AND EVALUATION FOR CRITICALITY SAFETY APPLICATIONS L. C. LEAL, K. H. GUBER, R. R. SPENCER, H. DERRIEN, AND R. Q. WRIGHT Oak Ridge National Laboratory Oak Ridge, Tennessee 37831 The Defense Nuclear Facility Safety Board (DNFSB) Recommendation 93-2 motivated the US Department of Energy (DOE) to develop a comprehensive criticality safety program to maintain and to predict the criticality of systems throughout the DOE complex. To implement the response to the DNFSB Recommendation 93-2, a Nuclear Criticality Safety Program (NCSP) was created including the following tasks: Critical Experiments, Criticality Benchmarks, Training, Analytical Methods, and Nuclear Data. The Nuclear Data portion of the NCSP consists of a variety of differential measurements performed at the Oak Ridge Electron Linear Accelerator (ORELA) at the Oak Ridge National Laboratory (ORNL), data analysis and evaluation using the generalized leastsquares fitting code SAMMY in the resolved, unresolved, and high energy ranges, and the development and benchmark testing of complete evaluations for a nuclide for inclusion into the Evaluated Nuclear Data File (ENDF/B). This paper outlines the work performed at ORNL to measure, evaluate, and test the nuclear data for aluminum for applications in criticality safety problems.
1
Data Measurements
The need for improved cross section data for aluminum was identified earlier in calculations performed by Palmer [1]. The calculations demonstrated that the infinite multiplication factors, k<,ff, for a mixture of 27A1/235U could change by as much as 10 to 12% relative to the ENDF/B, version VI, calculated value. Aluminum is an important material frequently present in reactor designs as part of the nuclear fuel. In addition, the US DOE is responsible for managing spent nuclear fuel for domestic and foreign research reactors. Therefore, there is a need for accurate cross sections for aluminum. To improve the data, neutron total and capture cross section measurements for aluminum were made at ORELA. ORELA is a high power white neutron source with an excellent time resolution that permits one to obtain accurate cross section measurements. The measurements for aluminum were made in the energy range from ~ 100 eV to several hundred kilo-electron-volts. The capture measurements were made using the time-of-flight technique with two rectangular aluminum samples of thicknesses 0.01520 at/b and 0.04573 at/b, respectively. The aluminum samples and the C6D6 detectors were located at a distance of 40 meters from the neutron target. The capture detector system was improved [2] to minimize the effect of structural material surrounding the sample and detectors to reduce the prompt neutron sensitivity. A 0.5mm thick 6Li-glass scintillator served as the neutron flux monitor. The pulse-height weighting was employed with the C6D6 detectors, and normalization of the capture
115
116 efficiency was carried out in a separate measurement using the black resonance technique by means of the 4.9 eV resonance from a 0.00508-cm-thick gold sample [3]. For the transmission measurements, two high purity samples with thicknesses of 0.0189 at/b and 0.1513 at/b, respectively, were used. The samples were mounted in a sampler changer at a location about 10 meters from the neutron-production target. The neutron detector was an 11.1-cm diameter, 1.25-cm thick 6Li-glass scintillator viewed on edge by two 12.7-cm diameter photomultiplier and positioned in the beam at a distance of 79.815 meters from the neutron source. Additional measurements were made to determine the gamma-ray background from the neutron source by putting a thick polyethylene sample in the neutron beam path. This was done by using a thick polyethylene sample located at 5 meters from the neutron target, which scattered all neutrons out of the beam at 80 meter fligh-path.3 2
Data Evaluation
A resonance evaluation of the aluminum transmission measurement and capture crosssection measurements was performed in the energy region from 0 to 845 keV. The evaluation used the SAMMY [4] code that uses the reduced R-matrix Reich-Moore formalism. In addition to the experimental data taken at ORELA, seven additional differential data sets were also used. The data summarized in Table 1. There are two transmission measurements taken at ORELA, two transmission measurements taken at GELINA (Geel Linear Accelerator, Geel, Belgium), [5] one capture cross section measurement taken at ORELA, and one transmission measurement done by Perey et al. at ORELA [6]. The total, scattering, and capture thermal cross sections values given in the literature were used to determine the cross sections at low energy [7].
Figure 1. Comparison of the ORNL transmission (left) and capture (right) with theoretical calculations. The resonance parameters obtained from the SAMMY analysis of the experimental data in the energy from thermal to 845 keV contain 79 resonances from which 63 resonances are in the energy range analyzed (14 s-wave, 29 p-wave, 20 d-wave), and 16 are external resonances that are used to account for the infinite series of resonances outside the evaluation range. In addition to the search for resonance parameters that
117 best fit the experimental data, SAMMY was also used to search for experimental effects such as background, normalization, sample effective temperatures, etc. This was done to assure that the uncertainties in the experimental data and the resonance analysis predict realistic uncertainties and covariance data. Table 1: Experimental Data Base
Author Guber et al.
Range Analyzed (eV) 0.5 to 4.0x105
Guber et al.
100 to 6.7x105
Rohr et al.
2.0xl0 5 to850xl0 5
Perey et al.
2.0xl0 5 to850xl0 5
Main Features Transmission (ORELA); TOF* 80 m; Samples 0.0189 and 0.1513 at/b Capture (ORELA); TOF 40 m; Sample 0.0457 at/b Transmission (GEEL); TOF 400 m; Samples 0.0533 and 0.1920 at/b Transmission (ORNL); TOF 47.35 m; Sample 0.7639 at/b
*TOF - Time-of-Flight Fig. 1 shows comparisons of transmissions and capture cross sections calculated from the resonance parameters (solid curves) with data (points) taken at ORELA. To our knowledge, this is the first time that the capture cross-section was used in the evaluation of the aluminum resonance parameters. The use of the capture cross-section helps to obtain realistic average capture widths. Fig. 2 shows a comparison of the transmission data of Rohr et al. (thickness 0.1920 at/b) and the transmission data of Perey et al. (thickness 0.7639 at/b) with the theoretical calculations in the energy region from 200 to 500 keV. The results shown in Figs. 1 and 2 indicate that the resonance parameter obtained from the SAMMY evaluation provides an excellent representation of the experimental data. In addition to the graphical representation, the average values of the total cross section and capture cross section are calculated using the resonance parameter and compared with the experimental values in Table 2. The average cross sections are calculated: Ea
= fc(E)dE Eb This is equivalent to a neutron microscopic reaction rate with a constant neutron flux. The results shown in Table 2 indicate that the calculated average cross sections are very close to the experimental values. The reaction rate is an important quantity in reactor calculations since it relates directly to the effective multiplication. Therefore, it
118
A
11 i\w^-\/
J
j
\ / v l v——nf *
^=c^r
Figure 2. Comparison of the transmission data of Rohr et al. (upper curve) and Perey et al. flower curve! with theoretical calculations from 200 eV to 500 keV.
is essential that the average neutron cross sections generated with resonance parameters reproduce the experimental results well. The experimental and calculated values of the cross sections at thermal (0.0253 eV) are compiled shown in Table 3. Table 2: Comparison of the average experimental and calculated cross sections
50 - 100 100-200 200 - 300
(Measured) (beV) 299.94 573.62 418.25
atE (Theory) (beV) 293.32 569.26 435.05
(Measured) (beV) 8.31xl0'2 0.161 0.982
(Theory) (beV) 8.85xl0"2 0.172 1.101
300 - 400
339.40
343.51
5.92xl0'2
6.55xl0"2
AE (keV)
CT,E
a
(",T) E
Table 3: The Cross Sections at Thermal (0.0253 eV) for 300 K.
Total Capture Scattering
Experimental 1.672 1.440 0.2321
ORNL 1.685 1.452 0.2334
ENDF/B-VI 1.606 1.374 0.2320
119
0.2
0,1
0.5
1.0
•.!>
2.0
Figure 3. Nearest neighbor-spacing distribution compared to a Wigner distribution.
3
Distribution of the Resonance Parameters
A great deal of knowledge about the structure of the nucleus can be obtained by examining the statistical properties of the resonance parameters. For instance, it is well known that the level spacing of adjacent resonances follows a Wigner distribution, whereas the reduced neutron widths follows a Porter-Thomas distribution, that is a % distribution with one degree of freedom. The level spacing distribution for .v-wave resonances, for which the total angular momentum are J=2 and J=3, in the energy region from thermal up to 850 keV is shown in Fig. 3. The average spacing fors-wave resonances is (60.0 ±6.1) keV. Fig. 4 shows the Dyson and Mehta A3-statistical test of the resonance parameters. A3 is a measure of the mean-square deviation between the number of observed energy levels and the fit to the number of levels in a straight line as a function of energy. For n energy levels, the A3 theory predicts that = [ln(«)=-.068]/;t2, with the variance given as KA3 = 1.169/7T4.
120
T h e 5-wave calculated value for is 1.89, whereas the expected value is 2.67 ±
1.09. Fig. 4 also shows the Porter-Thomas distribution of the s-wave resonances parameters in the energy range up to 850 keV. The average s-wave reduced neutron width calculated with the resonance parameters is (23.5 + 5.6) eV. The value of the swave strength function is (2.0 ± 0.5) x 10"*. The/7-wave strength function is (2.4 ± 1.0) x 10 . The value of the/?-wave strength function reported in the literature [6] is (2.6 ± 1.0) xlO"1.
Figure 4. Cumulative number of energy levels vs. energy for s-wave resonances in the energy region from thermal to 850 keV (left). Neutron-width distribution compared to a Porter-Thomas distribution (right).
4
Spin Channel Issue
In the process of performing the aluminum evaluation it was discovered that the ENDF format was not adequate for explicitly including the spin channel in the Reich-Moore representation. In the ENDF/B format there was no provision to represent more the one spin channel and the usual procedure is to combine all channels into one channel. This practice can lead to a distortion in the total cross section as shown in Fig. 5 for the resonance around the energy 778.5 keV. A new format has been approved by the Cross Section Evaluation Working Group (CSEWG) and has been implemented in the crosssection processing codes such as AMPX, NJOY, and MC2to circumvent this problem. 5
Data Testing of the Aluminum Evaluation
To test the efficiency of the resonance parameters for criticality safety applications related to reactors, a fictitious benchmark problem consisting of a critical sphere with a
121 mixture of 27A/235U was used. This problem is similar to that proposed by Palmer, which was an infinite medium problem. Palmer pointed out that better cross section data for aluminum were needed.
Figure 5. The channel spin effect in the total cross section for the resonance around 778.5 keV. The solid line is the cross section calculated with the complete spin channel representation; the dashed line is the one-channel representation.
The NJOY code system was used to generate 199-group (VITAMIN-B6) cross sections for the benchmark calculations. We calculated the infinite medium problem suggested by Palmer using the new evaluation and ENDF/B-VI aluminum evaluation. The keff computed with the new evaluation is 1.1043 whereas the ENDF/B-VI is 1.0071. Calculations were performed using the CSASN sequence, BON AMI, NIT AWL, and XSDRNPM, of the SCALE code system [8]. Table 4 shows the values of the 27A/235U number density ratios, the radius of the sphere, and the calculated kefr values using the new evaluation and ENDF/B-VI evaluation. The parameter EALF in Table 4 is the average lethargy of neutrons causing fission and indicates the "hardness" of the neutron spectrum. Table 4. Comparison of the k,n using the new aluminum evaluation and the ENDF/B-VI evaluation 27
A/235U
Radius (cm)
10 30 50 100 300 500 1000
42.497 61.605 79.093 165.2 260.2 308.5 392.25
199-group (ENDF/B-VI) 1.0128 1.0212 1.0310 1.0641 1.0276 1.0014 0.9484
New ORNL Evaluation 1.0115 1.0143 1.0191 1.0446 1.0431 1.0417 1.0280
EALF (keV) 353.24 207.02 139.00 27.96 3.81 1.39 0.35
122 6
Conclusions
A R-matrix resonance analysis of the experimental transmission data, total cross section, and capture cross section was performed for aluminum. A set of resonance parameters for the energy range from thermal to 850 keV was obtained. Comparison of the experimental and computed average cross sections shows good agreement. Benchmark calculations were compared using the new cross section library and the ENDF/B-VI evaluation. These results indicate huge discrepancies in kcffi mainly for systems with low EALF thermal systems using the existing cross section evaluation. Because the new evaluation reproduces experimental data and average cross section values, we recommend that this new evaluation replace the existing aluminum evaluation. Reference 1. "kcfr for Certain Metals Mixed with 235U," Criticality Safety Quarter, Sponsored by DOE-62, (Winter 1993). 2. P. E. Koehler, etal, Phys. Rev. C54, 1463 (1996). 3. K. H. Guber et al., "Neutron Capture and Neutron Total Cross Sections Measurements for 27A1 at the Oak Ridge Electron Linear Accelerator," Presented at 10th International Symposium Capture Gamma-Ray Spectroscopy and Related Topics, Los Alamos, New Mexico, August 30-September 3, 1999. 4. N. M. Larson, Updated Users' Guide SAMMY: Multilevel R-Matrix Fits to Neutron Data Using Bayes Equations, ORNL/TM-9179/R4 (December 1998). 5. G. Rohr et al., "Resonance Parameters for 27Al+n From Very High Resolution Transmission Measurements," International Conference on Nuclear Data for Science and Technology, Gatlinburg, Tennessee, May 9-13, 1994. 6. F. Perey et al., Test of the Neutron Total Cross Section Evaluation from 0.2 to 20 MeVfor C, O, Al, Si, Ca, Fe, andSi02, ORNL-4823 (1972). 7. S. F. Mughabghabefa/., Neutron Cross Sections - Neutron Resonance Parameters and Thermal Cross Sections, National Nuclear Data Center, Brookhaven National Laboratory, Upton, New York, 1981. 8. SCALE: A Modular System for Performing Standardized Computer Analysis for Licensing Evaluation, NUREG/CR-0200, Rev. 6 (ORNL/NUREG/CR/CSD-2R6), Vols. I, II, and III (May 2000). Available from Radiation Safety Information Computational Center at Oak Ridge National Laboratory as CCC-545.
NUCLEAR PHYSICS INVESTIGATIONS AT THE TIME-OF-FLIGHT SPECTROMETER GNEIS WITH SPALLATION NEUTRON SOURCE O.A. SHCHERBAKOV, A.B. LAPTEV, A.S. VOROBYEV Petersburg Nuclear Physics Institute, Gatchina, Leningrad district, 188300, Russia E-mail: laptev(a),pnpi.spb.ru
1
Facility description
The Gatchina neutron spectrometer GNEIS [1] with spallation neutron source is intended for neutron-nucleus interaction studies utilizing the time-of-flight technique over wide range of neutron energies from ~ 10"2 eV up to hundreds MeV. The spectrometer is based on the 1 GeV PNPI proton synchrocyclotron that has been used for physics experiments since 1975. A water-cooled lead target and polyethylene moderator are placed inside the vacuum chamber of the accelerator. A fast neutron burst results when the proton beam strikes this lead target. Five neutron beams are transported outside the vacuum chamber through the shielding wall of the machine main room in the laboratory building. Neutron beam lines 1-4 view the moderator, whereas beam line 5 views the lead target. Fig. 1 shows the general layout of the GNEIS facility. The high quality of the GNEIS's neutron source (high neutron intensity and short pulse) makes experimental capabilities of GNEIS comparable with that of the TOF-facilities at electron Linacs, ex. - ORELA (OakRidge), GELINA (Geel), and LUE-40/IBR-30 (Dubna), as well as high intensity proton accelerators, ex. - LANSCE (Los Alamos), and the recently developed n_TOF Collaboration (CERN). The main parameters of accelerator and spectrometer GNEIS are as follows: Pulsed neutron source: • average fast neutron intensity • duration of the fast neutron pulse • repetition rate • internal water-cooled rectangular lead target • rectangular polyethylene moderator Spectrometer: • • •
number of evacuated flight paths length of flight paths experimental area (GNEIS building)
123
~ 3-1014 n/s ~10ns < 50 Hz 40 cm x 20 cm x 5 cm 30 cm x 10 cm x 5 cm 5 35-50 m 45 x 30 m2
124 Physics equipment: 6
•
neutron detectors
• •
}*ray detectors fission fragment detectors
Li-glass, 3He-chamber, 47t-liquid scintillator tanks Nal(Tl), C6D6-liquid scintillators multiplate ionization chambers
The inserts in Fig. 1 show short titles of experiments performed at the GNEIS facility [2,3]- Some recent experiments have been carried out under support of Russian Foundation for Basic Research and International Science and Technology Center, in collaboration with research groups of other world-famous research centers. THE In.7/1 REACTION lll(,H I'KLCISION MF.ASCKEMENTS OK NEUTRON TOTAL CROSS SECTIONS: EVALUATION OF THE ELECTRIC PULARIZABILJTY OK THK NEUTRON IN I HE COULOMB FIELD OF HEAVY NUCLEUS
STL'BY OF THE NEUTRON CAITtlKF. REACTION FOR U-13S: MEASUREMENTS OFCT„,IN THE F.NERGY RANGE E„<10OKcV AND GAMMA-RAYS SPECTRA FROM CAPTURE OF RESONANCE NEUTRONS
STUDY OF THE NEUTRON INDUCED FISSION IN THE ENERGY RANGE L'P TO 10(1 M*V : MEASUREMENTS OF FISSION CROSS SECTIONS AND THEIR RATIOS, PUI.SK-HEICH1 SPECTRA OF FISSION FRAGMENTS
.COLLIMATORS
MEASUREMENT OF THE •'FORWARDBACKWARD" ASYMMETRY IN SLOW NEUTRON FISSION OF U-233 AND U-235: SEARCH FOR P-RESONANCE5 AND STUDY OF THEIR PROPERTIES
NEUTRON BEAMS POLYETHYLENE MODERATOR
Figure 1. General layout of the Gatchina neutron time-of-flight spectrometer GNEIS.
2
Study of the («,^-reaction in Neutron Resonances of 235U and 239Pu
Experimental studies [4] of the two-step («, ^-reaction gives unique information not only about fission process itself, but also about the structure of highly excited states in heavy nuclei, both in 1-st and 2-nd wells of the fission barrier, and radiative transitions between them. The fission y-ray multiplicity has been measured in neutron resonances of 235U and 239Pu. The experimental prefission widths r yf have been obtained from the observed correlations between the multiplicity of fission /-rays and reciprocal fission width Tfl of resonances: 235 U: r 7f (4") = 0.32meV±0.13meV 239Pu: r T f (l + )=1.9meV±0.8meV TV (3") = 0.87 meV ± 0.89 meV r Yf (0 + ) = 2.8meV±9.2meV Experimental and calculated prefission widths r yf are shown in Fig. 2 for the 4"resonances of 235U and l+-resonances of 239Pu as functions of the ratio of the £1 and Ml components in the prefission y-ray spectrum [4,5]. This comparison of the
125
experimental and calculated r yr widths shows the predominance of Ml radiation in the compound nucleus U and E\ radiation in the prefission spectra of y-transitions between the highly excited states. The best agreement between experiment and calculations was obtained by using models with intermediate damping of the vibrational states in the second well, and the GDR model. Ib\
lla'Vla
™
lib ' : '-~ * " • -
V
\
N\
U
\
\ . 1*-Resonances-
Ic
""ilib""--
E
lllc •
"Vv
v \
\
\ V"-
lllllllllll ll'llllllllllllllltlllliml
H
m
r,(Ei)/r,(Mi)
Figure 2. Experimental and calculated widths Tyf for 1+-resonances of 239Pu and 4"-resonances of 235U. Calculation model: I - single-humped fission barrier, II,III - double-humped barrier (complete and intermediate damping of the vibrational states in the second well, respectively); a - single-particle model (Weisskopf) for probabilities of y-transitions; b,c - GDR model (Axel-Brink, Lorentzian-shaped probability of partial y-transitions proportional to Ey4 and £,', respectively).
In another experiment at the GNEIS [4,6], the pulse-height spectra of fission y-rays have been measured for isolated resonances of 239Pu in the energy range from 10 eV to 91 eV. The difference pulse-height spectra for weak (Tf < lOmeV) and strong (r f > lOmeV) 1+resonances show a few structures that could be interpreted as prefission ytransitions between the levels at excitation energy 1-3 MeV below the neutron binding energy B„ (Fig. 3). 3
Pulse height MeV
Figure 3. The difference pulse-height y-rays spectra for weak and strong 1+-resonances of 239 Pu. Designations are the same as in a Fig. 2.
Measurements of the capture cross-section of U in energy range En<100 keV and y-ray spectra from the capture of resonance neutrons: study of the nature of the 721.6 eV resonance
The two lowest energy resonance clusters in the subthreshold fission cross-section of 238 U are dominated by resonances at 721.6 eV and 1211.4 eV. The anomalously small capture width of the 721.6-eV resonance (~ 4.7 meV) is strong evidence that this resonance is not a typical (class-I, corresponding the first well of fission barrier) compound state. If the 721.6 eV resonance is predominantly class-II (corresponding
126
to the second well) in character, then not only should its radiative width Vy be small, but its capture y-ray spectrum should be softer than that of other ,v-wave resonances (class-I). Prior to our measurements, Browne [7] observed a much softer y-ray spectrum for the 721.6 eV resonance than for neighboring resonances, whereas Weigmann et al. [8] found no difference. To resolve this contradiction, the capture y-ray spectra in isolated neutron resonances of 238U in the energy range from 400 eV to 1300 eV have been measured at the GNEIS [9]. The data were processed using a slightly modified version of the method of Weigmann et al. The idea was to detect a y-decay branch within the second well using two different bias values for the y-ray registration: lower B\ and upper 62. Then, for a value of 62 larger than 6„- En (2 MeV, height of the second minimum), the ratio R of resonance area Ay measured with two biases B\ and 62: R = A/bias B2)/Ay(bias B\) should be smaller for the resonance having major class-II fraction than for ordinary class-I resonances because the softer class-II component will be under the upper bias B2 for this resonance. The results of the present measurements [9] and those of Weigmann et al. , I [8] are shown in Fig. 4. As it is seen from : , ( : ^**'H our data, the capture y-ray spectrum of the 721.6 eV resonance is much softer than ''fri, that of the neighboring ,?-wave 1 resonances. Our data enable to make a conclusion that the 721.6 eV resonance is predominantly class-II by nature. As for the 1211.4 eV resonance, both our data Figure 4. Results of the capture y-ray and the results of [8] show that there are measurements for resonances of 238U. no solid arguments to consider this resonance as a class-II state. GEEL
GATCHINA
121,.,*
721.6 BV
•
Lower b e s B I :
1211.4 OV
•I
o
- 0 6 MeV
fewer bias B1 - 0 6 MeV
»
- 1 2 MeV
lippar bias B2 = 2.4 MeV
E,< E „
UPPER BIAS B2 (MeV)
4
Measurement of the neutron total cross-sections of Pb-isotopes: estimate of the electric polarizability of the neutron
The electric polarizability a„ is one of the characteristics of the neutron as an elementary particle and determines the induced electric dipole moment in an external electric field: d = a„ • E. It is known [10] that information about polarizability of the neutron can be obtained from the precise neutron total cross section measurements for heavy nuclei, for example 208Pb. Total cross section measurements of 208Pb and C from 1 eV to 20 keV have been performed at the GNEIS facility for the purpose of estimating a„ [11]. The total cross section measurements for carbon were performed as a null-test of the experiment. To take precisely into account the presence of other lead isotopes in the
127 sample, the total cross sections of 204Pb, 206Pb and 207Pb also were measured at GNEIS facility. A detailed description of the experiment can be found in [12]. The method employed for evaluation of a„ is described in detail in [13]. Fitting results for 208Pb and C are presented in Fig. 5. Points are experimental total cross section after the Schwinger and solid-state corrections and contribution of radiative absorption having been subtracted. The reduced tf in case of 208Pb is equal 2.5. The neutron polarizability obtained is a„ = (2.4+1.1)-10"3 fm3 and the amplitude of neutron-electron interaction is a„e = - (1.78±0.25)10 3 fm . The value of reduced tf in case of C is equal 1.7 . To eliminate influence of the distortions caused by uncertainties of experimental background the difference a(208Pb) - 2.42 -o(C) have been used for the neutron polarizability estimation. Fitting results for this difference are shown in fig. 6. Using this method, essentially the same polarizability was obtained, a„ = (2.44+1.32) -10"3 fm3, demonstrating the stability of the result. The value obtained for the amplitude of the neutron-electron interaction was ane- - (1.75±0.27)-10"3 fm, with a reduced y of 0.7.
10 10°
10 1
10:
10 3
10 4
Neutron energy, e V
Figure 5. Fitting results for 208Pb and 12C.
5
10
Neutron energy, eV
Figure 6. Fitting results for difference
a(mPb)-2.42a(C).
Measurement of the forward-backward asymmetry in slow neutron fission In the early 1980s, research on various parity-breaking effects and discussions of possibilities for searching for T-noninvariance effects near weak neutron p resonances in complex nuclei [14,15] attracted particular interest to the properties of these resonances. For heavy fissionable nuclei, however, essentially nothing was known at that time about the parameters and decay properties of low-lying (E„ « 1 keV) neutron p resonances that were not observed in the cross section. A new method to obtain such information is the study of the neutron energy dependence of the forward-backward asymmetry of the angular distribution of fission fragments [16] that results from s- and p-wave interference in neutron capture: W(G) = l + ofb(p„ -pj
128
where p„ and pf are the neutron and light fragment momenta. The measurements of the forward-backward asymmetry coefficient afa for 235U from 1 eV to 136 eV have been performed at the GNEIS. The results obtained for 235U in the energy range from 1 eV to 21 eV are shown in Fig. 7. Several irregularities caused by p-resonances have been observed in the energy dependence of the coefficient ofa. Estimations of the main /^-resonance parameters have been made. Fitting the data revealed that the average total width of the presonances was somewhat greater than that of j-resonances. For example in case of 235 U = (200 ± 50) meV and = (140 ± 10) meV. The information obtained in these measurements is very important for the fundamental investigations of the P- and ^-parity violation effects that are expected to be enhanced in a vicinity of presonances. The investigations initiated at the 8 10 12 14 16 GNEIS have been continued at JINR in Neutron energy, eV the course of collaborative research Figure 7. Energy dependence of the asymmetry carried out jointly with the Frank coefficient ap, and fission yield for 235U. Laboratory [17]. 6
Neutron induced fission cross-sections measurement of actinides and non-fissile nuclei relative to 235U in the energy range 1 - 200 MeV
During the last decade, measurements of neutron-induced fission cross-sections for some long-lived actinides and non-fissile nuclei have been systematically performed at the GNEIS facility. During 1997 to 2000, measurements were supported by 1STC grant # 609-97. Fission cross-section ratios have been obtained over a wide energy range of incident neutrons from 1 MeV to 200 MeV with special emphasis on energies above 20 MeV where previous experimental data were very scarce or absent. Measurements were performed simultaneously for each of the isotopic targets using a multi-plate ionization chamber and time-of-flight techniques on a 47m flight path. To date, fission cross-sections have been measured for 233U, 238U, 232Th, 239Pu, nat 209 'Np, Pb and Bi At present, other nuclides are under investigation, namely: Pu, Am (actinide) and W (non-fissile nucleus) with support from a new ISTC project (# 1971). Typical experimental data obtained for 238U and 239Pu (Fig. 8) [18], and for na,Pb and 209Bi (Fig. 9) [19] are compared with data of other authors, evaluations and theoretical calculations.
129 Jr^";-" ' '
'jji 3M
/^ '
"*
* *
""'^^H^i/ •
Pmsenl dala t_-
# F
•
1 1 1 1 •
l-.e
SKffi
7W
^n
Maslov
. .
i
Neutron energy, MeV
Neutron energy, MeV
Figure 8. Fission cross-section of 238U and 239Pu in the energy range 1 - 200 MeV.
:::::: ::: :::: ::: •
'^rfV~~\
""'Pb i
"il
::::::::::::!:::::
jr
k-
• o a o a
Present dala Smimov 1996 Rout 1950 Staples 1995 Smlmov 1999 Prokofiev 1996 ENDRHE-VI
II 100 Neutron energy, MeV
!
;;;;;;;:;;::;;;;;:;;:;!;;:::::;::::::::::|::::;::::::: = : p p g j £ S p i 1
••
}-••!•••;
209
Bi '"::::T::x:x::.'"i
..:.::::::::::::::::;:::::::::::.):::.':::
••i :::;~:::;-:;;:::::::::|::::: E!;;E:
:::::!:::::::: ! / ! ! :::::!:::::::: ff/ 1 ! !
• • ° • o
Present dala GokJanskl 195S Smlmov 1996 Elsmont 1996 Staples 1995 Smlmov 1997 ENDRHE-VI
II
P
I
100 Neutron energy, MeV
Figure 9. Fission cross-section of na'Pb and 209Bi in the energy range up to 200 MeV,
7
Acknowledgements
One of us (A.B.L.) thanks S. Raman and J.A. Harvey (ORNL, USA) for useful discussion. Various parts of the work described herein were performed in collaboration with G. A. Petrov, I. S. Guseva, Yu. A. Alexandrov, V. G. Nikolenko, I. L. Karpihin, P. A. Krupchitsky, A. M. Gagarskii, S. P. Golosovskaya, A. K. Petukhov, Yu. S. Pleva, V. E. Sokolov, A. Yu. Donets, A. V. Evdokimov, A. V. Fomichev, Yu. V. Toboltsev, T. Fukahori, A. Hasegawa, and V. M. Maslov. References 1. N.K. Abrosimov, G.Z. Borukhovich, A.B.Laptev etal., Nucl. Instr. Meth. A242(1985)p. 121. 2. O.A. Shcherbakov, Proc. of the VI Int. School on Neutron Physics, Alushta, October 8-18, 1990. (Dubna, JINR Report D3, 14-91-154, 1991) vol. 1, p. 98. 3. O.A. Shcherbakov, A.B. Laptev, G.A. Petrov, A.S. Vorobyev, Workshop on the Origin of the Heavy Elements: Astrophysical Models and Experimental Challenges, Santa Fe, New Mexico, September 3-4, 1999. (Los Alamos, LANL Report LA-13686-C, 2000).
130 4. O.A. Shcherbakov, Fiz. Elem. Chast. i atomn. yadra, v. 21, No. 2 (1990) p. 419. Translation: Sov. J. Part. Nucl. 21(2) (1990) p. 177. 5. O.A. Shcherbakov, A.B. Laptev, Proc. of 10-th Int. Symp. on Capture GammaRay Spectroscopy and Related Topics, Santa Fe, Aug.30 - Sept.3, 1999. Ed. by S. Wender. (Melville, New York, AIP, 2000) vol. 529, p. 710. 6. G.Z. Borukhovich, T.K. Zvezdkina, A.B.Laptev etal., In book "Neutron Physics. Proc. 6-th All Union Conf. on Neutron Physics, Kiev, October 2-6, 1983". (Moscow, CNIlatominform, 1984) vol. 1, p. 304. 7. J.C. Browne, Proc. Int. Conf. on the Interaction of Neutrons with Nuclei, Univ. of Lowell, Lowell, Mass., July 6-9. (Lowell, 1976) vol. 11, p. 1402. 8. H. Weigmann, G. Rohr, T. Van der Veen, G. Vanpraet, Proc. 2-nd Int. Symp. on Neutron Capture Gamma-Rays Spectroscopy and Related Topics, Petten, Netherlands, Sept. 2-6, 1974. (Petten, RCN, 1975) p. 673. 9. A.B. Laptev, O.A. Shcherbakov, LNPI Preprint 1712. (Leningrad, LNPI, 1991). 10. Yu.A. Alexandrov, Yad. Fiz. 37 (1983) p. 253; 38 (1983) p. 1100. 11. I.S. Guseva, A.B.Laptev, G.A. Petrov, O.A. Shcherbakov, Proc. of 10-th Int. Symp. on Capture Gamma-Ray Spectroscopy and Related Topics, Santa Fe, Aug.30 - Sept.3, 1999. Ed. by S. Wender. (Melville, New York, AIP, 2000) vol. 529, p. 713. 12. I.S. Guseva, A.B. Laptev, G.A. Petrov etal., PNPI Preprint NP-55-1999 2340. (Gatchina, PNPI, 1999). 13. I.S. Guseva, PNPI Preprint NP-27-1994 1969. (Gatchina, PNPI, 1994). 14. O.P. Sushkov, V.V. Flambaum, Usp. Fiz. Nauk 136 (1982) p. 3. Translation: Sov. Phys. Usp. 25 (1982) p. 1. 15. V.E. Bunakov, V.P. Gudkov, Nucl. Phys. A 401 (1983) p. 93. 16. A.M. Gagarskii, S.P. Golosovskaya, A.B. Laptev etal., Pis'ma Zh. Eksp. Teor. Fiz. 54, No.l (1991) p. 9. Translation: JETPLett. 54, No.l (1982) p. 7. 17. V.P.Alfimenkov, A.N.Chernikov, L.Lason et ai, Nucl.Phys. A645 (1999) p. 31. 18. O.A. Shcherbakov, A.Yu. Donets, A.V. Evdokimov etal., Proc. of IX Int. Seminar on Interaction of Neutrons with Nuclei "Neutron Spectroscopy, Nuclear Structure, Related Topics", ISINN-9, Dubna, May 23-26, 2001. (Dubna, JINR, E3-2001-192, 2001) p. 257. 19. O.A. Shcherbakov, A.Yu. Donets, A.V. Evdokimov el al., ibid. p. 326.
MEASUREMENTS OF NEUTRON CAPTURE CROSS SECTIONS OF LONG-LIVED FISSION PRODUCTS H. HARADA, S. NAKAMURA, K. FURUTAKA, T. KATOH, M.M. H. MIAH, AND 0 . SHCHERBAKOV Japan Nuclear Cycle Development Institute, Tokai-mura, Ibaraki 119-1194, E-mail: harada(a),tokai. inc. go. ;'p
JAPAN
H. YAMANA, T. FUJII, AND K. KOBAYASHI Research Reactor Institute, Kyoto University, Kumatori-cho, Osaka 590-0494, E-mail: vamana(3).HL.rri.kvoto-u.ac.ip
JAPAN
JNC has organized, together with Japanese universities, several projects on nuclear data measurements for long-lived fission products (LLFPs) and minor actinides (MAs) relevant to nuclear transmutation. The activation method, prompt gamma-ray spectroscopic method, and TOF method have been developed to determine the neutron-capture cross sections for important LLFPs and MAs from a thermal energy region up to a few MeV. Recent progress on these developments is reviewed and future data needs are also discussed.
1
Introduction
Neutron cross sections of long-lived fission products (LLFP) are important quantities as fundamental data for the study of nuclear transmutation of radioactive wastes. Especially, those of 79Se, 99Tc, 93Zr, ,07Pd, 126Sn, 129I, and 135Cs are very important since their half-lives are extremely long. The data for Sr and Cs are also important because their decay heat is so large. However, their neutron crosssection data are very limited and the improvement of the data accuracy is an important issue1'. For example, there are no experimental data on the neutron cross sections for 79 Se and l26Sn in the EXFOR data library. For 93Zr and 107Pd, there are no data on the thermal neutron capture cross sections. Although fast neutron cross sections are available for 99Tc, 93Zr, 107Pd, and 135Cs, the amount of data is very limited except the case of 99Tc. Even when several independent data sets are available as in the case of the resonance integrals for 99Tc, each result does not give a value consistent with the weighted mean value as is shown in Fig. 1. For the measurements of thermal neutron capture cross sections of Se, Zr, and 107Pd, the activation method that was used for the measurements of 99Tc2), 129I3), and 135Cs4) is not applicable since their neutron capture products are stable nuclei. The prompt gamma-ray spectroscopic method, therefore, has been developed for the measurements of these cross sections.
131
132
The energy dependence of the neutron capture cross section is also important to study a reactor system suitable for the transmutation of LLFP. However, the data is not enough for the use of feasibility study of the system. To supply systematic data from the thermal energy region up to a few MeV, JNC has organized, together with Japanese universities, several projects on nuclear data measurements for long-lived fission products (LLFPs) relevant to nuclear transmutation using the activation method, prompt gamma-ray spectroscopic method, and TOF method. Recent progress on these developments is reviewed in section 2, and future data needs are discussed in section 3. 30 •
ccpi
28 • 26
r PNC
i• 1
24 •
i *
22 •
1
Ui
£ 300 c
20 •
0)
18 • 16 •-
: T
« 400 -
ORNt
g 200 -
''
CD
14 -
« 100 tc
12 10
— i
1
1950
1960
1970
1980
—
i
1990
— 2000
HAR
i
« "1
Year
19S0
1960
1970
1980
Year
Figure 1. Thermal-neutron-capture cross section (left) and resonance integral (right) for function of publication year. Previous name of JNC is PNC.
2 2.1
99
Tc as a
Development of neutron-capture cross section measurement methods Recent activities on Activation methods
The neutron-capture cross sections of 90Sr(n, Y) 91Sr, "Tc(n,Y) 100 Tc, 137Cs(n,Y) 138m Cs and 166mHo(n, y)l67Ho reactions have been recently studied using the activation technique. The experimental data have been analyzed and the thermal neutron-capture cross sections and resonance integrals of 90Sr5) and 166mHo6'7) were deduced. The resonance integrals for these isotopes were determined for the first time in this study as 104± 16 mb and 10.0±2.7 kb, respectively. To obtain a precise neutron capture cross section for 137Cs, the production probability of isomeric state has been measured8' and it was used for the correction of the neutron capture cross section. In order to reduce the experimental error of the 99Tc(n,Y)100Tc cross section, the Y-ray emission probabilities of ,00Ru produced via the decay of 100Tc were also deduced ' using |3-y coincidence system10'. Figure 2 shows the photograph of the P~Y coincidence system. The error of the y-ray emission probabilities of 100Ru was
133
reduced to half of the previously reported errors11'. Table 1 shows the comparison of the results. Table 1. Comparison on gamma-ray emission probability of 591-keV transition from
~~ Detectors y-emission probability
Our data [9]
G. Berzins [11]
G e & Plastic 5.5±0.4%
N a l & Plastic 5.7±1.1%
100
Ru.
-it
W •\2t:
Figure 2. A photograph of the fb—y coincidence system made of a plastic scintillation counter and a high-purity Ge detector.
2.2
Prompt gamma-ray spectroscopy
The prompt y-ray spectroscopic method is a powerful tool for the measurements of thermal neutron capture cross sections. The method is applicable even when the activation method cannot be applied. The system for the prompt y-ray spectroscopy was newly developed and tested. The high-purity natZr and natPd foils were used as the irradiation targets. The irradiations were performed at B-4 neutron guide facility in Kyoto university reactor, wherein neutron flux was 5xl0 7 n/cm2/s. The prompt y-rays have been measured using two Ge detectors surrounded by BGO anti-coincidence detectors (Fig. 3) in y-y coincidence mode. To obtain a large photo-peak efficiency, two large Ge detectors (relative efficiency of each crystal was
134
90 % at 1.33 MeV) are used. The energy resolution at FWHM of each Ge detector is 2.3-2.5 keV at the energy of 1.33 MeV. To improve their peak-to-background (= total - peak) ratio, the Ge detectors are surrounded by anti-coincidence spectrometers made of BGO crystals of 50 mm in thickness. The recent study of large size Ge crystals for high-energy photons has shown their high peak-detection efficiency and peak-to-total ratio for high-energy photons. These characteristics will be advantageous for assigning y-rays in level schemes and determining the gamma-ray intensities. Once the y -ray intensities are determined, the thermal neutron-capture cross sections can be obtained from the single-mode measurement.
Figure 3. A photograph of the prompt gamma-ray spectroscopic system made of two high-purity Ge detectors surrounded by BGO anti-coincidence detectors at the beam line of thermal neutron guide in Kyoto university reactor.
2.3
TOF method
The neutron time-of-flight method enables one to obtain detailed information about the energy dependence of the neutron-induced reactions from the thermal energy region up to a few MeV, which is especially important for the study of transmutation systems based on a fast reactor. With an aim to develop the equipment and technique for neutron capture cross section measurements of FP and MA, capture cross section measurements utilizing the 16-section BGO y-ray detector (Fig. 4) have been carried out at the KURRI Linac. The neutron flux and energy resolution as well as the backgrounds at the 24 m flight path have been studied with the use of Au, Cd, C, Al, Bi, Pb, ' B samples and a set of the "black resonance" filters: Co,
135 Mn, Cd and In. Besides the conventional data acquisition system utilizing the MCS and ADC, a new measuring system based on the 40 MHz FLASH-ADC has been tested. Since the detector surrounds a sample with almost 4n geometry and the detection efficiency is high, this detector is suitable when only small amount of sample is available. If the system developed will be used in an intense pulsed neutron facility, this advantage can be further enlarged.
Figure 4. A photograph of the 16-section BGO y-ray detector
3
Discussion
Activation method is very powerful for the measurement of neutron-capture cross sections of radioactive samples for thermal neutrons, where small amount of sample is required. When this method is applied, however, one needs precise data on emission probabilities of gamma rays emitted from neutron-capture products as was shown for the case of Tc in section 2 . 1 . For some of the fission products as 79Se, 107Pd, and 93Zr, the activation method is not applicable for the measurement of thermal-neutron-capture cross sections since their neutron-capture products are stable. In spite of the importance, there are almost no data for these isotopes. The prompt y-ray spectroscopic method described in section 2.2 is one of the solutions, although this method requires tremendous efforts on the data analysis. If the intense pulsed neutron source becomes available, the transmission measurement using time-of-flight method will be also applied to deduce the neutron capture cross section for these radioactive samples. The crosscheck of the data measured using different methods will contribute to the improvement of the data accuracy.
136 The preparation of radioactive sample is also important issues since the contamination in the sample severely affects the results on neutron cross sections. JNC, together with Tokyo Institute of Technology and Kyoto University, started a new program for preparation of 79Se, ' 7Pd, and 3Zr, including the collaboration with ORNL. The carefully prepared samples will be used for the measurement of neutron capture cross sections from thermal to MeV energy region. 4
Acknowledgements
This work has been carried out in part under the Visiting Researcher's Program of the Research Reactor Institute, Kyoto University. O. S. acknowledges support by JNC as an International Fellow. M. M. H. M. is grateful to STA Fellowship of Japan. References 1. Furutaka K. et al., Nuclear Data Measurements for P&T and Future Plans in JNC, Proceedings of the 6lh OECD/NEA Information Exchange Meeting on Partitioning and Transmutation of Actinides and Fission Products, Madrid (Spain), 11-13 December 2000, EUR 19783 EN, OECD/NEA, Paris (France), 2001. 2. Harada H. et al., Measurement of Thermal Neutron Cross Section and Resonance Integral of the Reaction 99Tc(n, Y)10°Tc, J. Nucl. Sci. Technol. 32 (1995)pp.395^03. 3. Nakamura S. et al., Measurement of Neutron Capture Cross Section and Resonance Integral of the l29I(n, y)l30I Reaction, J. Nucl. Sci. Technol. 33 (1996)pp.283-289. 4. Katoh T. et al., Measurement of Thermal Neutron Capture Cross Section and Resonance Integral of the Reaction 135Cs(n, y)'36Cs, J. Nucl. Sci. Technol. 34 (1997)pp.431-«8. 5. Nakamura S. et al., Measurement of the Thermal Neutron Capture Cross Section and the Resonance Integral of the 90Sr(n, y)91Sr Reaction, J. Nucl. Sci. Technol. 38 (2001) pp.1029-1034. 6. Harada H. et al., Measurement of Effective Neutron Capture Cross Section of 166m Ho using Two Step Irradiation Technique, J. Nucl. Sci. Technol. 37 (2000) pp.821-823. 7. Katoh T. et al., Measurement of Thermal Neutron Capture Cross Section and Resonance Integral of the 166mHo(n, y)I67Ho Reaction using a Two-Step Irradiation Technique, to be published in J. Nucl. Sci. Technol. 39 [7] (2002).
137 8.
Wada H. et al., Production of the Isomeric State of Cs in the Thermal Neutron Capture Reaction 137Cs(n, y)'38Cs, J. Nucl. Sci. Technol. 37 (2000) pp.827-831. 9. Furutaka K. et al., Precise Measurement of Gamma-ray Emission Probabilities of 100Ru, J. Nucl. Sci. Technol. 38 (2001) pp.1035-1042. 10. Furutaka K. et al., Evaluation of R-y Coincidence Measurement System Using Plastic Scintillation (3-ray Detector Developed for the determination of y-ray Emission Probabilities of Short-lived Nuclides, J. Nucl. Sci. Technol. 37 (2000) pp.832-839. 11. Berzins G., Bunker M. E. and Starner J. W., Energy Levels of 100Ru, Phys. Rev. C187(1969)pp.l618-1632.
T E M P E R A T U R E M E A S U R E M E N T S IN D Y N A M I C A L L Y - L O A D E D S Y S T E M S USING N E U T R O N R E S O N A N C E S P E C T R O S C O P Y (NRS) A T LANSCE
V. W. YUAN Los Alamos National Laboratory, Los Alamos, NM USA E-mail: [email protected] In previous attempts to determine the internal temperature in systems subjected to dynamic loading, experimenters have usually relied on surface-based optical techniques that are often hampered by insufficient information regarding the emissivity of the surfaces under study. Neutron Resonance Spectroscopy (NRS) is a technique that uses Doppler-broadened neutron resonances to measure internal temperatures in dynamically-loaded samples. NRS has developed its own target-moderator assembly to provide single pulses with an order of magnitude higher brightness than the Lujan production target. The resonance line shapes from which temperature information is extracted are also influenced by non-temperaturedependent broadening from the moderator and detector phosphorescence. Dynamic NRS experiments have been performed to measure the temperature in a silver sheet jet and behind the passage of a shock wave in molybdenum.
1
Introduction
Measurement of the internal temperature of a system undergoing dynamic loading is of great interest in many areas of physics and chemistry. Existing methods of temperature measurement in dynamic systems have various inherent limitations. Optical methods rely on viewing the sample surface, and during dynamic events the surface may become obscured from view by reaction products. Even if the surface can be clearly viewed, model calculations are required to relate temperatures at the surface to those present in the interior. Sensors such as thermocouples placed in the interior of the sample can be inactivated, destroyed or altered during the dynamic event. Furthermore, the presence of invasive diagnostics change the dynamics of the very interaction one desires to study. The use of Doppler broadening in neutron resonances as a quantitative way to measure temperatures has been proposed [1,2] and investigated for cases of static or quasi-static temperature measurements. Neutrons are probes that can interact throughout the volume of a sample to detect the temperature in its interior. When neutrons pass through a sample material, the time-of-flight (TOF) spectrum of the transmitted neutrons exhibits a series of dips or resonances characteristic of the material in the beam. These resonances appear when neutrons are captured from the beam in the formation of excited states in the A+l nucleus ( n+A -> (A+l)* ). Subsequent de-excitation of these states, by gamma emission or particle emission into An steradians, effectively eliminates these neutrons from the transmitted beam.
138
139 The resonance locations and line shapes which appear in the TOF spectrum are unique to each isotopic element, and temperature determinations can be localized through the positioning of the appropriate isotope as a resonant tag in isolated regions of the sample. 2
Neutron Resonance Spectroscopy
In Neutron Resonance Spectroscopy (NRS), the temperature of an irradiated sample is determined from the line shape of a resonance. Resonances in the epithermal energy range between 1 and 100 eV are typically used. The energy available when neutron interacts with a nucleus is dependent on the relative velocity. As a result, the observed line shape is a convolution between a Lorentzian and a Gaussian with a width a dependent on both temperature and neutron energy: o =2(EkT/A)1/2. One must also be aware, however, that other factors that do not depend on temperature may also influence the resonance line shape. First, since the moderation of fast neutrons to epithermal energies is produced by a series of random collisions in the moderator, not all neutrons of a given final energy will leave the moderator at the same time. Second, scintillation detectors such as 6Li-loaded glass can produce phosphorescent tails in their light emission that will affect the resonance line shape. And third, motion of the sample's center of mass can result in a shift in the location of a resonance's centroid. Hence, velocity distributions within the sample can produce the superposition of a range of shifts, resulting in a distortion to the line shape. 3
Neutron Beam and Experimental Setup
At the Los Alamos Neutron Science Center (LANSCE), most neutron experiments are performed using the production target of the Manual Lujan Neutron Scattering Center (MLNSC). Protons accelerated by the 800-MeV LINAC are accumulated in a proton storage ring (PSR) and then released as short (125 ns), intense (2.2 x 1013 protons) pulses at a 20-Hz repetition rate. In normal operation, these pulses strike a tungsten production target at MLNSC and produce fast neutrons that are then moderated in water and serve as a source to several time-of-flight (TOF) beam lines. The neutron source for NRS measurements is a separate 2 U/CH2 target/moderator assembly located in the "Blue Room" at LANSCE. When this assembly is installed, single PSR proton pulses are delivered to the "Blue Room" for the production of neutrons. One advantage of the "Blue Room" neutron source is that samples under study can be placed in much closer proximity (within 1 meter) to the source than is possible at MLNSC (> 6 meters) because of the bulk shielding required for the latter to protect users from radiation. The time resolution in an NRS
140
temperature determination is directly proportional to the proximity of the sample to the neutron source. The radiation shielding of the "Blue Room," however, is not designed to handle the same intense neutron production rates as is the shielding at MLNSC, and therefore the delivery of single PSR pulses is restricted to a maximum delivery rate of one pulse every 5 minutes. The present beam transport systems also do not support simultaneous beam delivery to both MLNSC and the "Blue Room," and hence NRS must presently operate in a "sole-use" beam mode that is incompatible with other beam lines. A schematic of the apparatus used in the NRS measurements is shown in Fig. 1. An intense, short-duration PSR pulse of 800MeV protons (125ns fwhm, 3 x 1013) is incident on a 238U target, and the resulting high-energy neutrons are moderated to epithermal energies in a block of CH2. The moderating CH2 is set in a wing-view arrangement so that the detector's view of the moderator does not include a direct view of the 238U target, thereby reducing high-intensity radiation from the target that could directly enter the detector. A reflector comprising of beryllium blocks surrounds the target/moderator to increase neutron flux. Neutrons transmitted by the sample are collimated and subsequently detected by a 6Li-glass/phototube array positioned 23m from the neutron source. In order to fully utilize the high-intensity neutron flux, we sample the neutron signal in a transient digitizer (current mode detection) [3] rather than attempt to count individual neutrons. A current loop monitors the arrival of the protons at the neutron production target. The timing signal from the accelerator is used to accurately initiate the dynamic events being studied. For geometries with low count rates, where individual neutron pulses are detected without overlap, we have the option to discriminate and count individual neutron pulses in a multiscaler (pulse-counting mode) so as to eliminate the contribution of phosphorescence in the Lithium glass.
125 ns
timin
9 Pickoff
236 u target
high flux - > take data in "current-mode"
^•r
CH 2 epithermal neutrons
containment vessel and sample
°l_i-glass detector array
Figure 1. Schematic of NRS experimental setup
141 4
Measurement Time Scale
The transit time for resonant-energy neutrons to pass through a sample being studied determines the time scale of the temperature measurement. In order for the measurement to be meaningful, the thermodynamic state of the sample must remain constant during the passage of the resonant neutrons. In other words, the more rapid the dynamic event to be studied the shorter the required transit time. Neutron resonances typically have natural widths of order 100 meV. In a TOF measurement, an energy spread AE will, at the sample, translate into a time spread directly proportional to the distance between the sample and the neutron source: At =
1 tAE 2
LAE
where L = distance from source to the sample
E
As an example of the time scales involved, Fig. 2 shows the resonance profile for the 21.1-eV resonance in l82W. The FWHM of the resonance is 170 ns for a sample 1 meter from the source. As can be seen, it is possible to study dynamic systems on a sub-microsecond time scale.
900 -
21.1-eV resonance in 1 8 2 w|
I
BOO -
^ 700 -
600 -
500 -
170 ns - 4
IW*I^WA
V-
.J.IAI*J»AWI.,U K L
\/
\[ time (ps)
Figure 2. Line shape showing the transit time for neutrons in the 21.1 -eV resonance in
5
W.
Contributions to the Final Line shape
In neutron-nucleus interactions, the basic Lorentzian line shape of an isolated neutron resonance is modified by the motion of target nuclei because the energy available to the interaction depends on the relative neutron-nucleus velocity. If one assumes that the target nuclei have a Maxwellian velocity distribution, then the resulting Doppler-broadened cross section has the form of a convolution of a Lorentzian and Gaussian (Voight profile):
142
CT(v)= \a<\v-V\)P{V)dV=-Zj=
P-Vl
where
A
-TTdE r/2 J
HmEkT
1
°°ee
4,!
and then cr(E) = a., T= ^-dy " 2t4n _{ \+y
A
where
t=— r
In addition, the observed line shape can possess contributions from other, non temperature-dependent sources. One is the energy-dependent background that is due to gamma rays and non-resonant neutrons that have their source in room-return radiation and fast neutrons moderated in the material around the beam line between sample and detector. We have measured room return neutrons using detectors located out of the neutron beam and found the room-return backgrounds to be small. A second contribution to the resonance line shape is the time-dispersion properties of the neutron target/moderator and surrounding reflectors. Monte Carlo calculations using the computer programs MCNP and LAHET have been performed to determine the time characteristics of the moderator dispersion. The dispersion calculations show the presence of two (0.4/VE US and 1.5/\E us) time-dispersion components, where E is given in eV. A third contribution to the resonance line shape is the presence of phosphorescent tails in the light emitted by the lithium-glass scintillators of the detector array. To determine the time dependence of these tails a radioactive Californium source and a digital scope were used to studied the line shape of timeaveraged detector pulses. To study the longer time constant components, a program was written to record single photoelectron events in multiscaler mode. The analysis shows the presence of phosphorescence components with time constants varying from 100ns to several hundred us. Other contributions to the resonance line shape include the pulse width (125ns fwhm) of the PSR proton pulse and the filter constant used to match the detector pulse to the sampling interval of the transient digitizer. The approximate relative sizes of components contributing to the NRS line shape are shown as a function of neutron energy in Fig. 3.
143
Time Widths at 23 m
widths at 23m — intrinsic — Doppler(T=1000K) Doppler (T=298 K) --- moderationl --- moderation2 — PSR pulse e delta_t_int
2
3
4 S 6 7 B 9
1
Z
3
4 S 6 7 B 9
10
E(eV) Figure 3. Curves showing various width contributions to resonance line shapes as a function of neutron energy. The curve labeled intrinsic approximates the trend of resonance intrinsic widths with energy. The points that are labeled ' d e l t a t i n t ' show actual widths for resonances in tungsten.
6
Shocked Metal Experiment
NRS can be used to determine temperatures in a wide range of dynamicallychanging systems. Systems whose temperatures are of interest include: materials through which a Shockwave has passed, explosives after the passage of a detonation wave, frictionally heated interfaces, and metal jets. The design behind the experiment to measure the temperature in a shocked metal is shown in Fig. 4. An aluminum flyer plate is explosively launched to a velocity of 3.5 km/s and strikes a block of molybdenum. Upon impact, a shock wave that traverses the molybdenum is generated. Molybdenum was chosen for a target because it is a standard for numerous other shock tests. Inside the block of molybdenum is a localized region doped with 1.7 atomic percent l82W. The 21.1-eV resonance in 182W is used to restrict the temperature measurement to the doped region only. In this way, with proper timing the temperature inside the sample can be measured immediately after the passage of a Shockwave, but prior to the entrance of rarefaction waves approximately 1 microsecond later. Fig. 5 shows resonance data taken before and after passage of the shock. In comparison to the resonance line shape for the unshocked molybdenum, the shocked molybdenum line shape is seen to be both broadened and shifted in time. The large difference in overall neutron flux between
144
the shocked and unshocked case results from explosive products that highly attenuate the neutron flux. Preliminary analysis indicates the shocked molybdenum to be at a temperature of between 750 and 900 degrees K, depending on the pressure generated by the Shockwave. Flyer (Al) v>3.5 km/s 182W-doped — Molybdenum (1.7 atm %)
" I 6mm
Shockwave front (630 kbar)
rarefaction waves Figure 4. Illustration showing the design of the experiment to measure the temperature in a metal after passage of a Shockwave. The 1 microsecond time window for the measurement begins with the passage of the shock and ends with the return of rarefaction waves from surface of the molybdenum target.
before shock after shock
21.1 eV Tungsten 182
1/
^^^
^"^^^ilf^^iN^^
V 360
T 380
Time (microseconds) Figure 5. Comparison of the 21.1 eV resonance in W before passage of a shock with the same resonance after passage of the shock. The resonance is broadened and the centroid of the resonance is shifted.
145 Acknowledgements I wish to thank my other NRS collaborators without whose hard work and contributions the results described in this paper could not have been achieved. The collaboration is between several Los Alamos National Laboratory Divisions and groups: David Funk and Ron Rabie from DX-2, Larry Hull from DX-3, David Bowman and George Morgan from P-23. References 1. P.H. Fowler and A.D. Taylor, Temperature Imaging Using Epithermal Neutrons', in Neutron Resonance Radiography, Report of a workshop held at Los Alamos National Laboratory, LA-11393-C (1987) pp. 46-80. 2. J. Mayers, G. Baciocco and A.C. Hannon, Temperature Measurement by Neutron Resonance Radiography, Nucl. Instrum. & Meth. A275, 453 (1989) pp. 453-459. 3. J.D. Bowman, J.J. Szymanski, CD. Bowman, V.W. Yuan, A. Silverman, and X. Zhu, Current Mode Detector for Neutron Time-of-Flight Studies, Nucl. Instrum. & Meth, A297, 183 (1990) pp. 183-189.
RADIOACTIVE TARGET PRODUCTION AT RIA J. C. BLACKMON Physics Division, Oak Ridge National Lab, P.O. Box 2008, Oak Ridge, TN 37831 US E-mail: [email protected] We explore the production of samples of long-lived isotopes (tm > 1 h) at an advanced radioactive ion beam facility, RIA. Production yields at RIA are compared to capabilities at stable beam facilities and at high-flux reactors. Long-lived neutron-rich nuclei can generally be produced more efficiently in a nuclear reactor if appropriate target samples are available. As a result, only two ,s process branch point nuclei, l35Cs and 163Ho, seem suitable for sample production at RIA. In contrast, samples of many long-lived proton-rich nuclei are produced effectively at RIA, including isotopes important for the p process. Sample production at RIA is more favored when the lifetime of the isotope is shorter.
1
Introduction
Our understanding of the origins and structure of the matter that makes up our world has been shaped by decades of study of the atomic nucleus. Radioactive nuclei have played an important role in this basic research and in the many practical applications that have resulted. However, the scope of nuclear physics has been limited mostly to the study of isotopes with combinations of protons and neutrons similar to those of stable nuclei. The development of intense beams of radioactive ions has begun to provide a powerful tool for producing and studying exotic isotopes. The compelling scientific program with such beams has been outlined previously in several works (see for example [1,11,15]). The opportunities for research with radioactive ion beams has led to a proposed concept for an advanced radioactive ion beam facility, the Rare Isotope Accelerator (RIA) [14]. The RIA facility is now advocated by the nuclear physics community in the U.S. as the highest priority for a major new research facility [9]. A primary goal of RIA is to produce beams of exotic nuclei with short halflives, but RIA will also have the capability to produce very intense beams of longlived radioactive isotopes. Isotopes with half-lives of hours or longer can often be studied more efficiently by implanting into a sample for study at a later time, for example as a target for bombardment by a neutron beam or stable ion beam. In some cases, such as in nuclear decay studies or medical research, it may be advantageous or necessary to study samples using equipment or facilities not available at RIA. In this work we examine the possibilities for production of samples of isotopes with t|/2 > 1 h using radioactive ion beams at RIA. One primary motivation for the study of some long-lived radioactive nuclei is their important role in the synthesis of heavy elements. We focus in particular on the
146
147 production of samples important for the astrophysical s process and p process, but the production of a variety of other radioactive samples is also discussed. We assume in this work that the radioactive ion beam from RIA will be dedicated for producing a sample of a long-lived isotope by implantation for a period of about 1 day. Radioactive samples may be produced at RIA without using the primary radioactive ion beam. For example, undesired radioactive isotopes that are eliminated from the beam in a fragment separator might be collected for production of a long-lived sample. This would be advantageous since it would not require dedicated use of the RIA beam, for which time will be highly competitive. However, the feasibility of such techniques requires further study. 2
Brief Overview of the RIA facility
Radioactive ion beams have most often been produced using one of two approaches: the Isotope Separator On-Line (ISOL) technique [10] or projectile fragmentation (PF) [8]. In the ISOL technique, a light ion beam (typically hydrogen or helium) bombards a thick target to produce radioactive isotopes. The target is contained in a vacuum enclosure with an ion source, and some fraction of the reaction products diffuse out of the thick target and effuse to the ion source where they can then be ionized, extracted, and accelerated. The beam intensities produced via the ISOL technique depend critically on the chemical properties of the target material and the production isotope, and on the lifetime of the production isotope. While beams of some radioactive isotopes can be copiously produced using the ISOL technique, others are produced weakly or not at all. The PF technique produces a broad range of radioactive isotopes by bombarding a thin target with an energetic (E = 0.1-1.0 GeV/u) heavy ion beam. The radioactive ions emerge from the target with a high energy, are analyzed and focused by a mass spectrometer, and transported to the experimental area. The primary advantages of this approach are that it is independent of the chemistry of the elements involved and that beams of short-lived exotic nuclei may be produced. However, the high energies of PF beams are not suited for many classes of experiments, and the beam quality (purity and emittance) is worse than obtained using the ISOL technique. RIA is envisioned as a facility that combines standard and novel techniques for radioactive ion beam generation with a flexible, high-power driver accelerator [14]. In addition to standard ISOL and PF, one new approach to be applied at RIA involves stopping PF beams in a buffer gas, then extracting and reaccelerating them. Measurements with a prototype gas catcher have demonstrated that this process can be performed very efficiently in time scales of ~10 ms [13]. This technique combines some of the best features of ISOL and PF, allowing any element, including short-lived isotopes, to be produced with good beam quality. Another technique for
148 radioactive ion beam generation at RIA is "in-flight fission". This approach is similar to PF, except that an incident actinide beam is fissioned to produce neutronrich radioactive products. A variant of the ISOL technique to be used at RIA involves the use of neutrons to induce fission in an ISOL target. This approach produces a different distribution of isotopes and mitigates one of the technical challenges of ISOL beam production, dissipation of the heat generated in stopping the charged-particle beam in the thick target. A combination of all these techniques will be applied at RIA in order to produce a vast array of radioactive ion beams with unprecedented intensities. The intensities for different isotopes at RIA has been estimated by Jiang et al. using a combination of these different approaches [3]. We have primarily used these projected intensities in our analysis of long-lived sample production at RIA. 3
Isotopes of interest for the s process
About half of the nuclei heavier than iron are synthesized in Asymptotic Giant Branch (AGB) stars by a series of neutron captures and beta decays called the .? process [7]. The synthesis of elements in the * process is reasonably well understood owing to precise measurements of neutron capture cross sections on a large number of nuclei and to the availability of good observational data. Some important questions remain, particularly regarding the conditions in the astrophysical environment where the s process occurs. Several nuclei along the s process path have half-lives on the order of a year or longer. These nuclei live sufficiently long that they may undergo neutron capture in the s process before decay. An accurate knowledge of the neutron capture cross sections for these "branch point" nuclei would help determine the thermodynamic conditions where the s process occurs, but the cross sections for most of these nuclei remain unmeasured or poorly known. Measurements may finally be possible for many of these nuclei using the intense neutron flux from the Spallation Neutron Source (SNS) currently under construction at Oak Ridge National Laboratory. The pulsed neutron flux from the SNS will exceed that of other facilities by more than a factor of 30, hence reducing the number of sample atoms required for a measurement of the cross section [6]. A list of nuclei that are important s process branch points is given in Table 1. Guber et al. have calculated the capture cross sections for these nuclei and estimated the minimum number of targets atoms required for a measurement of the neutron capture cross section at the SNS [2]. The number of sample atoms and time required to produce such a sample at RIA is shown in Table 1. Of the 17 s process isotopes considered, 6 (135Cs, 147Pm, 153Gd, 163Ho, 170Tm, and 171Tm) are expected to have sufficient intensities at RIA so that a sample could be produced in a reasonable length of time (~ days). However, some isotopes can be produced more abundantly by other means, especially when activation in a high-flux reactor is possible. A good
149 example is Pm which is currently planned for production at the ILL Reactor by activating a sample of neodymium [12]. The 146 Nd(«,y) 147 Nd reaction ( o = 1.4 b) produces Nd which decays to Pm with an 11 day half-life. A very pure sample of l 4 7 Pm can be produced by chemical separation since it is the only promethium isotope produced by activation of neodymium with a half-life longer than days. A sample of greater than 10 19 atoms of 147 Pm is expected to be produced in an a single activation, about 1000 times the quantity that could be produced in 3 days of running at RIA. Table 1. Estimated time needed for production of samples of s process branch point nuclei at RIA.
Isotope ^ ^ 85 Kr 94 Nb 106 Ru ,35 Cs ,47 Pm 155 Eu 153 Gd 163 Ho ,69 Er 170 Tm 17, Tm 179 Ta ,82Hf 185 ,93
W Pt
204 rp.
Half-life l.lxl0 6 y 10.7 y 2xl0 4 y 367 d 2xl0 6 y 2.62 y 4.9 y 241.6 d 4570 y 9.4 d 128.6 d 1.92 y 1.7y 9xl0 6 y 75.1 d 50 y 3.77 y
Atoms [2] (10'5) 50 200 20 70 50 9 5 4 4 20 4 10 8 40 20 7 80
Rate [3] (109pps) 20 80 1 5 3000 40 4 20 400 30 100 100 1 10 2 1 1
Time (days) 26 29 230 160 0.2 2.6 14 2.3 0.1 6 0.5 1.1 93 46 115 81 930
The l70Tm and 17lTm isotopes can also be produced abundantly by activation. The 170Tm isotope can be produced by 169Tm(w,y)170Tm with a cross section of 105 b. An irradiation of thulium (which does not require enrichment) at a high-flux reactor would produce a factor of 104 more l7 T m than could be produced in a day of running at RIA. Isotopic separation will be needed to separate the 16 Tm from the 170 Tm, but this added complication is certainly justified by the increased yields. The 171 Tm isotope can be produced via 170 Er(«,y) ,71 Er(P) 171 Tm with a cross section of 5.8 b. An enriched sample of ,70 Er (natural abundance 15%) would be needed, but about 10 3 more 171 Tm could be produced than would be generated in a day of beam time at RIA.
150 Of the 17 -v process branch point nuclei studied, only 135Cs and 163Ho can be produced competitively at RIA. These isotopes can be also be produced by 134 Xe(«,y)135Xe(p)135Cs and 163Dy(>,«)163Ho respectively. The cross sections for these methods of production are smaller than for those isotopes described previously, and the handling and chemical separation considerations are nontrivial. In contrast, RIA produces very intense beams of these isotopes, and satisfactory targets can be produced in less than 1 day of beam time. 4
Isotopes of interest for the p process
There are about 3 dozen stable heavy isotopes that cannot be produced by neutron capture from lighter elements. These neutron-deficient isotopes have a low natural abundance, and it is plausible that they have their origin in a common astrophysical process that has been named the p process [7]. The currently favored scenario involves the spallation of more abundant stable isotopes in an extreme astrophysical environment such as supernovae, but the process is not well understood. Many of the important reactions in the p process are (y,n) and (y,a) reactions on proton-rich radioactive nuclei. The rates for these reactions have to be predicted on the basis of theoretical models that have substantial uncertainties arising from the input optical model parameters. Measurements of («,oc) reactions on certain long-lived protonrich radioactive nuclei would provide an important constraint on the nuclear parameters in p process models. Koehler has estimated the («,a) cross sections for several important nuclei and calculated the sample size required for («,a) cross section measurements [5]. The minimum time required to produce such samples at RIA is shown in Table 2. In cases where the cross section has not been previously estimated, a minimum sample size of 10 atoms is assumed corresponding to an average total («,oc) cross section of ~ 1 b. Radioactive target production is feasible at RIA for about half of the p process isotopes considered. Several of these isotopes cannot be easily produced by other means. Production at RIA will be the only viable means to produce samples of 146 Sm, l48Gd, l50Gd, 154Dy, 172Hf and 194Hg. Most of the other isotopes can be produced by (p,n) or (d,n) reactions, but the production yields are typically significantly lower than can be produced at RIA. For example, Au can be produced via the l95Pt(p,n)l95Au reaction. Approximately 1015 atoms/day of 195Au would be produced by irradiating a 1-mm-thick disk of platinum with 20 uA of 30 MeV protons, as might be typically available at a small cyclotron facility. The anticipated production yield at RIA will be about 80 times larger than by the (p,n) reaction. The discrepancy between yields from the different production techniques is not as great in many cases. For example, the production rate for 145Pm at RIA is about an order of magnitude larger than that by obtained by proton irradiation of neodymium. However, proton-rich isotopes can generally be produced more
151 copiously at RIA than by other approaches if the anticipated RIA beam intensity is greater than 1 pnA. Table 2. The estimated beam time required for production of samples of radioactive nuclei of interest for the astrophysical p process at RIA. A required sample of 1016 atoms was estimated in cases (indicated by asterisks) where the cross section for the reaction of interest was very uncertain.
Isotope S3 Mn 55 Fe 57 Co 59 Ni 91 Nb 92 Nb 93 Mo 97 Tc 109 Cd 137 La ,33 Ba 139 Ce ,43 Pm 145 Pm 145 Sm 146 Sm ,48 Gd ,50 Gd ,54 Dy 159 Dy ,57Tb ,72Hf 195
Au ,94 Hg 202pb
Half-life 3.7xl0 6 y 2.73 y 272 d 7.6xl0 4 y 680 y 3.5xl0 7 y 4000 y 4xl06y 461 d 6xl04y 10.5 y 138 d
18y 1.7 y 340 d lxl08y 75 y 1.8xl0 6 y 3xl06y 144 d 71 y 1.9 y 186 d 444 y 52500y
Atoms [5] (IP 15 )
Rate [3] (10 9 pps)
200 200 200 100 40 20 40 20 8 14 20 20 9 6 6 25 10* 10* 10* 3 4 10* 5 10* 10*
2 2 2 2 2 2 3 3 500 10 80 20 200 100 100 200 150 300 1000 2000 1500
Time (days) 1000 1000 1000
500 220 110 150 75 0.2 16 3 12 0.5 0.7 0.7 1.4 0.7* 0.3* 0.1*
0.2 0.03
20
6*
1000
0.06
800 1
100*
1*
Table 3. The estimated size of some samples with ti/2 ~ 1 day that could be produced at RIA.
Isotope 64 Cu 67 Cu 72 Zn 66 Ga 72 Ga 71 As 73 Se 76 Br 77 Br 79 Kr 8, Rb 86 Rb 85 Sr 105 Rh U1 Ag ,07 Cd 125 Sn 13, I 135j 125
Xe Cs 136 Cs 140 Nd ,27
,53^ ,53
Dy Dy 166 Ho ,57
,67
Tm
166
Yb
.87jr 195
5
Hg
Half-life 12.7 h 61.8h 46.5 h 9.5 h 14.1 h 63.3 h 7.15 h 16.2 h 57.0 h 35.0 h 4.6 h 18.6 d 64.8 d 35.4 h 7.5 h 6.5 h 9.6 d 8.0 d 6.6 h 16.9 h 6.3 h 13.1 d 3.4 d 2.3 d 6.4 h 8.1 h 26.7 h 9.25 d 56.7 h 10.5 h 9.9 h
Rate [3] (10 9 pps) 30 10 20 40 20 100 30 900 1000 3000 4000 800 4000 10 3000 300 100 500 300 3000 8000 1000 20 400 400 400 100 900 3000 7000 800
Atoms (10' 5 ) 1 0.9 2 1 1 9 0.8 50 90 300 70 50 300 0.9 80 7 9 40 7 200 200 90 2 30 9 10 9 80 300 300 30
Nuclei with shorter half-lives
We have also examined a variety of other nuclei with half-lives of ~ 1 day for preparation as radioactive samples at RIA. A list of some potential isotopes, and the
153 estimated sample sizes that could be produced in less than 1 day of beam time at RIA, are shown in Table 3. Nuclei with this range of half-lives were selected in part because they are difficult to produce by other techniques since the half-life of the isotope limits irradiation time. In fact, some of these isotopes (67Cu, 72Zn, 73Se, 86Kr, 92 Sr, 113Ag, 135I, M0Nd, M9Tb, 153Dy, and l66Yb) are especially difficult to produce by other means. Nuclei with half-lives of hours to days are also potentially interesting for applied research since the nuclei can be produced and efficiently used in a diagnostic measurement in a short-period of time with no long term radiological effects [4]. For example, 73Se may be a potentially interesting isotope as a medical diagnostic. Selenium tracks certain biological functions in mammals, and Se could be used as a source for PET scans. 6
Conclusions
We have examined the feasibility for production of samples of long-lived radioactive isotopes (t1/2 > 1 h) using dedicated radioactive ion beams from the RIA facility. In general, we find that neutron-rich radioactive nuclei can be produced more efficiently at a high-flux reactor if an appropriate target sample for irradiation exists. Proton-rich radioactive nuclei can be produced competitively at RIA, especially nuclei with shorter half-lives. Other methods for production of long-lived samples at RIA deserve further investigation. 7
Acknowledgements
We thank P. E. Koehler and S. Raman for many interesting and insightful discussions. Oak Ridge National Laboratory is managed by UT-Battelle, LLC, for the U.S. Department of Energy under Contract No. DE-AC05-00OR22725. References 1. Gelletly W., Science with radioactive beams: the alchemist's dream. Cont. Physics 42 (2001) pp. 285-314. 2. Guber K. H., Koehler P. E., Valentine T. E., and Leal L. C, Neutron cross section measurements at the spallation neutron source. To be published in the Proceedings of the International Conference on Nuclear Data for Science and Technology, Tsukuba, Japan (2001). 3. Jiang C. L., et al., Ria Yields. http://www.phv.anl.gov/div/rib/ria yields/yields home.html (2000).
154 4. Karataglidis, S., RIA Applications Workshop Summary, http://www.la11l.g0v/oras/t/workshop/homepage.htm. (2000). 5. Koehler P. E., Neutron Nuclear Astrophysics at Spallation Neutron Sources. In Fundamental Physics with Pulsed Neutron Beams, ed. by Gould C. R., Greens G. L., Plasil F., and Snow W. M. (World Scientific, Singapore, 2001) p. 127. 6. Koehler P. E., Comparison of white neutron sources for nuclear astrophysics experiments using very small samples, Nucl. Instrum. Meth. Phys. Res. A460 (2001) pp. 352-361. 7. Meyer, B. S., The r-, s-, and p-processes in nucleosynthesis. Annu. Rev. Astron. Aslrophys. 32 (1994) 153-190. 8. Morrissey D. J. and Sherrill B. M., Radioactive nuclear beam facilities based on projectile fragmentation, Phil. Trans. Roy. Soc. (London) A356 (1998) p. 19852006. 9. Nuclear Science Advisory Committee, Symons J. P. (chair), Opportunities in Nuclear Science: A Long Range Plan for the Next Decade (2002). 10. Ravn H. L., Radioactive ion-beam projects based on the two-accelerator principle, Phil. Trans. Roy. Soc. (London) A356 (1998) p. 1955-1984. 11. RIA Steering Group, Casten R. F. and Nazarewicz W. (co-chairs), RIA Physics White Paper, http://www.phy.anl.gov/div/rib/ria-whitepaper-2000.pdf (2000). 12. Rundberg R. S., "Off-line" radioactive targets, these proceedings. 13. Savard G., Schwartz J., Caggiano J., Greene J. P., Heinz A., Maier M., Seweryniak D., and Zabransky B. J., ISOL beams from fragmentation: the best of both worlds. Nucl. Phys. A 701 (2002) pp. 292c-295c. 14. Savard G., The U.S. Rare Isotope Accelerator Project. In Proceedings of the 2001 Particle Accelerator Conference, ed. by Lucas P. and Webber S. (IEEE, Piscataway, 2001) pp. 561-565. 15. Smith M. S. and Rehm K. E., Nuclear Astrophysics Measurements with Radioactive Beams. Annu. Rev. Nucl. Part. Sci. 51 (2001) pp. 91-130.
P A R I T Y VIOLATION I N E P I T H E R M A L N E U T R O N RESONANCES G. E. M I T C H E L L North Triangle
Carolina State University, Raleigh NC 27695-8202, USA and Universities Nuclear Laboratory, Durham NC 27708-0308, USA J. D . B O W M A N A N D S. I. P E N T T I L A
Los Alamos
National
Laboratory,
Los Alamos
NM 87545,
USA
E. I. S H A R A P O V Joint
Institute
for Nuclear
Research,
Dubna
141980,
Russia
The TRIPLE Collaboration has studied parity violation in p-wave neutron resonances by measuring the cross section longitudinal asymmetries at neutron energies up to 300-2000 eV, depending on the target. The measurements were performed at the pulsed spallation neutron source of the Los Alamos Neutron Science Center. Parity violations were observed in 75 resonances of Br, Rh, Pd, Ag, Cd, In, Sn, I, Sb, Cs, Xe, La, U, and Th. Statistical methods were developed to determine the weak interaction rms matrix elements and the corresponding spreading widths r „ . The value of < rw > is about 1.8 X 1 0 - 7 eV. Although the individual weak spreading widths are globally consistent with a constant or slowly varying mass dependence, there is also evidence for local fluctuations.
1
Introduction
Following the discovery of very large parity violating effects for neutron pwave resonances by the Dubna group, 1 the Time Reversal Invariance and Parity at Low Energies (TRIPLE) Collaboration was formed to study parity violation in compound nuclei. The high neutron flux available at the Manuel Lujan Jr. Neutron Scattering Center (MLNSC) at the Los Alamos Neutron Science Center (LANSCE) was very well suited for these experiments. A statistical ansatz was adopted: the compound nucleus is considered to be a chaotic system and the symmetry-breaking matrix elements random variables. The final result of such a parity violation experiment is the root-mean-square symmetry-breaking matrix element, which is obtained from a set of measured longitudinal asymmetries {P]E for individual resonances. The crucial point is that the rms matrix element can be obtained without detailed information about the wave functions of the compound states. For a p-wave resonance at energy E, the asymmetry p is defined by a±(E) = ap(E)(l±p),
155
(1)
156
where o- ± (£) is the neutron cross section for the + and - neutron helicity states, and ap(E) is the p-wave resonance cross section for unpolarized neutrons. In practice the parity violation measurements are feasible only near a maximum of the p-wave neutron strength function. The initial TRIPLE measurements with 2 3 2 Th and 238 U were near the maximum of the 4p neutron strength function. In order to obtain information concerning possible mass dependence of the effective nucleon-nucleus weak interaction, our attention turned to the A = 110 mass region, where the 3p neutron strength function maximum is located. Due to the lower level densities in this mass region as compared with those in the region of the 4p maximum, the PNC effects for even-even nuclei were much smaller than observed for 2 3 2 Th or 238 U. Therefore we focused on odd mass targets with their higher level densities, and correspondingly larger PNC effects. Parity nonconserving (PNC) effects were observed for all but one of the many odd mass targets that we studied in this mass region. However, the analysis for targets with non-zero spin is complicated and requires the knowledge of spectroscopic information for the s- and p-wave resonances. Results from the early measurements are discussed in reviews by Bowman et al.2, Frankle et al.3 and Flambaum and Gribakin. 4 After the initial measurements we improved the experimental system, repeated and improved the early measurements, and carried out experiments with many additional targets. The most recent reviews are by Mitchell, Bowman, and Weidenmuller5 and by Mitchell et al.6 The latter provides a comprehensive summary of the TRIPLE experimental results and analysis methods. 2
Experimental Method
Measurements of the PNC asymmetries p were performed at the MLNSC pulsed neutron source7 at LANSCE. The apparatus developed by the TRIPLE Collaboration to measure p is described in a number of papers, including the original experimental layout 8 , the neutron monitor, 9 the polarizer, 10 the spin flipper,11 and the neutron detector. 12 The layout of the polarized neutron beam line for the present PNC experiments is given in Crawford et al.13 The measurements were performed on flight path 2, which views a gadoliniumpoisoned water moderator. After the moderator the neutrons are collimated to a 10-cm diameter beam inside a 4-m thick biological shield. The neutrons then pass through a 3 He/ 4 He ion chamber system 9 that acts as a flux monitor. Next, the neutrons pass through a polarized-proton spin filter.10 Neutrons with one of the two helicity states are preferentially scattered out of the beam, leaving a beam of longitudinally-polarized neutrons (with polarization / n ~
157
70%). Fast neutron spin reversal (every 10 s) was accomplished by passing the neutron beam through a spin reversal device consisting of a system of magnetic fields.11 The neutron spin direction was also changed by reversing the polarization direction of the proton spin filter approximately every two days. For most of the targets the PNC effects were measured by transmitting the neutron beam through samples located at the downstream part of the spin nipper. The 10 B-loaded liquid scintillation detector 12 was located 56.7 m from the neutron source. This 55-cell segmented detector can handle instantaneous counting rates up to 9 MHz per cell with a dead time of about 20 ns. The data acquisition process is initiated with each proton burst. The detector signals are linearly summed and filtered. An ADC transient recorder digitally samples the summed detector signal 8192 times in intervals determined by the filtering time. The 8192 words are added, as a 'pass', to a summation memory for 200 beam bursts before being stored. The data from 160 passes form a 30-minute 'run' for the data analysis. The shape analysis was performed with the code FITXS, 14 which was written specifically to analyze the time-of-fiight (TOF) spectra measured by the TRIPLE Collaboration. The multilevel, multichannel formalism of Reich and Moore 15 was used for the neutron cross sections, which were convoluted with the TOF resolution function described by Crawford et al.13 The neutron resonance energies and gT n -widths were obtained and then held fixed. The parameters {fnp)+ and (fnp)~ in equations a ± n = ap[\ + (/ n p)±],
(2)
were fit separately for the + and — helicity TOF spectra. Here o^, is the experimental neutron cross section for the + and — neutron helicity states and fn is the absolute value of the neutron beam polarization. The asymmetry p defined by Eq. (1) was calculated as „ P
3
=
[(/nP) + - (fnP)-} fn[2+{fnP)+ + (fnP)-Y
, . ^
Analysis
The analysis to obtain the weak matrix elements and spreading widths from the PNC cross section asymmetries is described by Bowman et al.16 The essential argument is that the observed PNC effect in the /U-th p-wave resonance is due to contributions from a number of neighboring s-wave resonances v. Since there are several mixing matrix elements and one measured asymmetry,
158 one cannot obtain the individual matrix elements V^v. If the weak matrix elements connecting the opposite parity states are random variables, then the longitudinal asymmetry is also a random variable. From the distribution of the asymmetries one can infer the variance M2 of the individual matrix elements — the mean square matrix element of the PNC interaction. The analysis relies on knowledge of the spectroscopic parameters. The essence of our approach to the likelihood analysis is to include all available spectroscopic information and to average over remaining unknowns. The net result is that more information reduces the uncertainty in the rms value of the weak matrix element. In general the observed PNC asymmetry for a given p-wave level /i is
^ ^ ^ V ^ ,
(4)
where <7Ml/2 and gv are the neutron decay amplitudes of levels ^ and v (g2 = 9li/2 +#M3/2 — r ^n a n d 9l = rV n ), Ey. and Ev are the corresponding resonance energies, and Vvli is the matrix element of the PNC interaction between levels fi and v. For targets with Iv = 0, the s-wave resonances have 1/2+ and the pwave resonances l / 2 ~ or 3/2~. Only the 1/2 resonances can mix and show parity violation. We assume that the values of the asymmetries measured for different p-wave resonances have mean zero and are statistically independent. The likelihood function for several resonances is therefore the product of their likelihood functions. The mean square matrix element M2 is the variance of the distribution of the individual PNC matrix elements VVI±. The quantity p^ in Eq. (4) is a sum of Gaussian random variables VVfl each multiplied by fixed coefficients, and is itself a Gaussian random variable. 17 The variance of pM is (A^Mj)2, where
'-yf^ ^^=\E^TjTt-
<5>
The probability density function of the longitudinal asymmetry p is
p(
exp
^>=^ (-4^)-
(6)
The analysis for I ^ 0 targets is much more complicated and is discussed at length in Bowman et al.11
159 4
Results
Table I lists the parity violations observed by the TRIPLE Collaboration and their relative sign. There is a non-statistical anomaly in 2 3 2 Th, where the 10 longitudinal asymmetries observed up to 250 eV all have the same sign.31 As suggested in Table I, there is evidence that this phenomenon is restricted to 2 3 2 Th. All attempts to explain the sign correlation as a general effect led to unreasonably large values of the weak single particle matrix element. The observation of PNC effects with the opposite sign at higher energies in 232 Th 3 2 support a local doorway explanation 33 with a doorway as a random statistical event.
Table 1. Parity violations observed by TRIPLE.
Target 81 Br 93 Nb 103 Rh
Reference
104pd
21
105pd 106pd 108pd
21,22
18 19 20 21
21,22
107
Ag Ag 113 Cd
23
109
23
H5In
25
117
Sn 121 Sb 123 Sb
26
127j
27
131
Xe Cs 139 La 232 Th below 250 eV 232 Th above 250 eV
28
133
29
238JJ
13
total total excluding Th
24
27 27
30 31 32
All 1 0 4 1 3 2 0 8 4 2 9 4 5 1 7 1 1 1 10 6 5 75 59
P+ 1 0 3 0 3 0 0 5 2 2
5 2 3 0 5 0 1 1 10 2 3 48 36
P0 0 1 1 0 2 0 3 2 0 4 2 2 1 2 1 0 0 0 4 2 27 23
160 One key question is whether the effective nucleon-nucleus weak interaction is mass dependent. The weak spreading widths are listed in Table II. The likelihood functions for the measured spreading widths are asymmetric, often with tails to large values of Tw. The likelihood functions for the logarithm of the weak spreading widths are much more symmetric. We performed a leastsquare analysis in terms of x = lnr^,. Our analysis is approximate because it assumes Gaussian distributions for x. The central limit theorem implies that the distribution of the average of x should have a probability density that tends towards a Gaussian as more data are included, even if the individual data have non-Gaussian probability density functions. We use the average of the upper and lower errors as an approximate symmetric error for x. We first calculated the average value of x and the value of \2 assuming that all x's are described by a common value. The reduced x2 is 2.46, which is not consistent with this hypothesis. The nuclide with the spreading width that gives the largest contributions to x2 is 133 Cs. If we remove 133 Cs, the reduced x2 *s acceptable - 0.96. Although the complete set of measured spreading widths is not consistent with one common value, the data set without 133 Cs and 93 Nb (which showed no statistically significant parity violation) can be described with a single value of Tw = 1.8lo'.3 x 10~ 7 e ^ For targets with 1 = 0, the matrix elements are obtained directly from the likelihood analysis in terms of Mj. For 1^0 nuclei, the matrix elements Mj are obtained from the spreading widths r„, using the level spacings Dj. Except for 93 Nb and 133 Cs, the values of the matrix elements are concentrated in a narrow range — 0.5 to 2.2 meV. 5
Conclusion
The TRIPLE Collaboration studied 20 different nuclides and measured the longitudinal asymmetries for several hundred p-wave resonances. Statistically significant PNC effects were found in 75 resonances. In the process a large number of new resonances were observed, particularly for the p-wave resonances. Either the weak matrix element or the weak spreading width was determined directly from the experimental longitudinal asymmetries. Although in general it is not possible to determine the individual weak matrix elements, it is possible to determine the rms weak matrix element (or spreading width) for a given nuclide. The rms weak matrix elements determined in this manner are qualitatively consistent with a variety of theoretical approaches. Calculations of the weak rms matrix elements by Rodin and Urin 34 and by Flambaum and Vorov35 give qualitative agreement with the measured matrix elements. The average weak spreading width T^, = 1.8 x 10~ 7 eV is qualitatively
161 Table 2. Weak interaction spreading widths.
A 93
Nb 103 Rh 104pd 105pd 106pd 107
Ag Ag 113 Cd 109
115Tn 117
Sn Sb 123 Sb
121
127T
133
Cs
232Th 238 TJ
/ 9/2 1/2 0 5/2 0 1/2 1/2 1/2 9/2 1/2 5/2 7/2 5/2 7/2 0 0
T™ (10- 7 eV) <0.1 1 4+ 1 ' 2
Reference 19 20
i-o^: 4 7 0.8±0;35 i.o^:87
21
2.7±?-i
23
q+2.5
23
q 0+3.4 °-z-1.6 1 q + 0.8
24
i
o.3±° 0 ;
21 21
25 26
6 2
4.8±|:S
27
1 9+ 0 6 +0 - 9 o no6+0018 4u.uuo_ 7+2-70 004 ^ • ' - 1 .180 1 3+ '
27
15
27 29 31 13
consistent with theoretical expectations, e.g., with a recent result for 238 U 36 obtained in the framework of the statistical spectroscopy approach. Globally the weak spreading widths are consistent with a constant value or with a slowly varying mass dependence. There is evidence for local fluctuations in the spreading widths — perhaps similar to those observed in isospin symmetry breaking. 37 A more detailed quantitative comparison, probably emphasizing the values obtained for different mass regions and the apparent anomalies for specific nuclei, requires additional theoretical effort.
Acknowlegments This work was supported in part by the U.S. Department of Energy, Office of High Energy and Nuclear Physics, under grant No. DE-FG02-97-ER41042, and by the U.S. Department of Energy, Office of Energy Research, under Contract No. W-7405-ENG-36.
162 References 1. V. P. Alfimenkov, S. B. Borzakov, Vo Van Thuan, Yu. D. Mareev, L. B. Pikelner, A. S. Khrykin, and E. I. Sharapov, Nucl. Phys. A 398, 93 (1983). 2. J. D. Bowman, G. T. Garvey, Mikkel B. Johnson, and G. E. Mitchell, Ann. Rev. Nucl. Part. Sci. 43, 829 (1993). 3. C. M. Frankle, S. J. Seestrom, N. R. Roberson, Yu. P. Popov, and E. I. Sharapov, Phys. Part. Nucl. 24, 401 (1993). 4. V. V. Flambaum and G. F. Gribakin, Prog. Part. Nucl. Phys. 35, 423 (1995). 5. G. E. Mitchell, J. D. Bowman, and H. A. Weidenmuller, Rev. Mod. Phys. 71, 445 (1999). 6. G. E. Mitchell, J. D. Bowman, S. I. Penttila, and E. I. Sharapov, Phys. Rep. 354, 157 (2001). 7. P. W. Lisowski, C. D. Bowman, G. J. Russell, and S. A. Wender, Nucl. Sci. Eng. 106, 208 (1990). 8. N. R. Roberson et al., Nucl. Instrum. Methods Phys. Research A 326, 549 (1993). 9. J. J. Szymanski et ai, Nucl. Instrum. Methods Phys. Research A 340, 564 (1994). 10. S. I. Penttila, J. D. Bowman, P. P. J. Delheij, C. M. Frankle, D. G. Haase, H. Postma, S. J. Seestrom, and Yi-Fen Yen, High Energy Spin Physics, edited by K. J. Heller and S. L. Smith (American Institute of Physics, New York, 1995), p. 532. 11. J. D. Bowman, S. I. Penttila, and W. B. Tippens, Nucl. Instrum. Methods Phys. Research A 369, 195 (1996). 12. Yi-Fen Yen et ai, Nucl. Instrum. Methods Phys. Research A 447, 476 (2000). 13. B. E. Crawford et ai, Phys. Rev. C 58, 1225 (1998). 14. J. D. Bowman, Y. Matsuda, Y.-F. Yen, and B. E. Crawford (unpublished) . 15. C. W. Reich and M. S. Moore, Phys. Rev. I l l , 929 (1958). 16. J. D. Bowman, L. Y. Lowie, G. E. Mitchell, E. I. Sharapov, and Yi-Fen Yen, Phys. Rev. C 53, 285 (1996). 17. W. T. Eadie, P. Drijard, F. E. James, M. Roos, and B. Sadoulet, Statistical Methods in Experimental Physics (North Holland, Amsterdam, 1971), p. 204. 18. C. M. Frankle et ai, Phys. Rev. C 46, 1542 (1992). 19. E. I. Sharapov et ai, Phys. Rev. C 59, 1131 (1999).
163 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.
D. A. Smith et al, Phys. Rev. C 60, 045503 (1999). D. A. Smith et al, Phys. Rev. C 65, 035503 (2002). B. E. Crawford et al, Phys. Rev. C 60, 055503 (1999). L. Y. Lowie et al, Phys. Rev. C 59, 1119 (1999). S. J. Seestrom et al., Phys. Rev. C 58, 2977 (1998). S. L. Stephenson et a/., Phys. Rev. C 61, 045501 (2000). D. A. Smith et al., Phys. Rev. C 64, 015502 (2001). Y. Matsuda, Phys. Rev. C 64, 015501 (2001). J. J. Szymanski et al, Phys. Rev. C 54, R2576 (1996). E. I. Sharapov et al., Phys. Rev. C 59, 1772 (1999). V. W. Yuan et al, Phys. Rev. C 44, 2187 (1991). S. L. Stephenson et al, Phys. Rev. C 58, 1236 (1998). E. I. Sharapov et al, Phys. Rev. C 6 1 , 025501 (2000). M. S. Hussein, A. K. Kerman, and C.-Y. Lin, Z. Phys. A 351, 30 (1995). V. A. Rodin and M. G. Urin, Phys. Part. Nuclei 31, 490 (2000). V. V. Flambaum and O. K. Vorov, Phys. Rev. Lett. 70, 4051 (1993). S. Tomsovic, Mikkel B. Johnson, A. Hayes, J. D. Bowman, Phys. Rev. C 62, 054607 (2000). 37. H. L. Harney, A. Richter; and H. A. Weidenmiiller, Rev. Mod. Phys. 58, 607 (1986).
NEW EXPERIMENTAL CAPABILITIES FOR PARITY NONCONSERVATION AND TIME REVERSAL INVARIANCE VIOLATION IN NEUTRON TRANSMISSION S. I. PENTTILA Los Alamos National Laboratory, Los Alamos NM 87545, USA E-mail: [email protected] Parity-violation studies in neutron resonances are firmly established. With the intense epithermal neutron beam of the SNS statistics of these measurements can be significantly improved. A combination of the SNS beam, optically-polarized 'He neutron beam polarizer/analyzer, and a polarized n 9 La target will allow for the first time to search for a violation of time-reversal invariance (77?/) in neutron transmission with a level of sensitivity where it would be of fundamental significance. We briefly discuss the size of the effect of the TRI violation in neutron transmission and describe recent technical developments towards PV experiments and a test of TRI in lowenergy neutron interactions.
1.
Introduction
The existence of a parity-violating (PV) time-reversal invariant (TRI) component in the N-N interactions is experimentally well established [1,2,3]. At low energies the PV N-N interaction is typically described in terms of a potential derived from a single meson exchange (n±,p,anda>) [4]. In this picture one of the interaction vertices is weak and the other one is strong. CP invariance rules out any exchange of neutral pseudoscalar mesons like n in these purely PV interactions. In addition to the exchanged meson, the PV weak meson-nucleon coupling constants are determined by the exchanged isospin. Though, a number of experiments have been performed in N-N or few-body systems, we currently have only upper limits for the weak coupling constants [1]. This slow pace is due to the small signals making the experiments very difficult to do. Most of the available observables depend on a combination of these couplings, and we do not have enough sensitive data sets to extract a value for an individual coupling. With the new high-precision data [5] and with results from experiments under construction [6], in a few years, the values with small uncertainties will be known. The characteristic size of the /'-odd T-even weak force has been estimated to be on order of 10"6 [1]. However, largely enhanced PV effects, up to a few percent, have been measured in transmission of polarized neutrons through unpolarized nuclear targets [7,3]. The P-violating observable in these measurements is a longitudinal asymmetry
<+(rr
~ (Es-Ep) Jrv
164
165
where c + (crZ) is the total p-wave cross section for the + (-) neutron helicity state, V,p„ is the nuclear weak matrix element of P-odd T-even interactions between s- and /?-wave resonances. The large asymmetries are understood resulting from a large wave function admixture by the weak force due to the small level spacing (Es - Ev » 10 eV) between the parity mixed resonances and a mixing of large s-wave neutron amplitudes ^Tm into smallp-wave amplitudes JT [3,8]. Most of the resonance parameters are available with a limited accuracy from nuclear spectroscopy. There is an increasing interest to search for a ^-violation in N-N interactions, especially, in the transmission of neutrons through a polarized target. Observation of a P-violation in the experiments would also imply CP-violation through the CPT theorem, and thus the T-violation experiments would help to understand the origin of the CP-violation. In analog with the treatment of P violation, a N-N interaction that simultaneously violates P and TRI at low energies can be described in terms of a N-N potential generated by the exchange of single light mesons [9]. In the P-violating interaction the exchange of neutral spin-zero mesons is not forbidden, and at least n° has to be included when the P- and P-violating N-N interactions are described. The scale of the strength of the P-odd T-odd effects in NN interactions has been estimated in a number of papers. At present, the most stringent bound comes from the experimental limit on the neutron electric dipole moment (EDM) dn [10]. The correlation dna-E, where a is the neutron spin and E an electric field, is a P- and ^-violating observable. From the EDM limit, the size of the P-odd T-odd couplings of about 10"11 is obtained [9,11-13]. Observables depending on these couplings are well beyond the reach of any experiment being considered at present, therefore, possible enhancements have been studied. The measured large enhancements in P-odd P-even asymmetries in neutron pwave resonances have led to a prediction of an enhanced sensitivity also in Tviolating effects in/?-wave resonances [11]. In transmission of neutrons through a target medium the total neutron cross section is proportional to the coherent forward scattering amplitude that is expressed by [14,15] (2) f(00) = A + BAdI+BLaB + Cdk + DdIxk + Ea(fk)(Ixk), where 6 and k are unit vectors of the neutron spin and momentum and / is the unit vector parallel to target spin. The coefficients A, BA, and BL represent the strong interaction part of the scattering amplitude. The coefficient A gives the spin independent absorption of the neutron flux. The real and imaginary parts of the coefficient BA give the rotation of the neutron spin about the direction of the target polarization, the Abragam rotation, and the spin dependent attenuation of the neutron flux, respectively. BL gives the Larmor precession. The real and imaginary parts of the coefficient C give the PV spin rotation about k and the PV longitudinal asymmetry of eq. (1). The real and imaginary parts of the coefficient D describe a possible simultaneous P- and P-violating asymmetry and a neutron spin rotation about the vector / x k. The term C in the amplitude/is P-odd P-even, whereas the
166
E term is P-even P-odd, thus being a test of a pure ^-violation. A search of the D correlation requires a polarized nuclear target, while a search of the E correlation uses an aligned target [16]. In a neutron transmission experiment, a neutron detector yield N(d,I,k) depends on the amplitude of eq. (2). The coefficients of the amplitude are selected by reversing appropriate vectors. According to the optical theorem, the difference of the total cross section of the P-odd P-even C term in the transmission of polarized neutrons is Aa/,=^Im(/+-/J = ^ImC,
(3)
K
K
where + (-) indicates the neutron helicity state. In practice, the imaginary part of C appears in a p-wave resonance as a small change in the detector transmission yield (4) N±=N0exp[-n(Jp(l±PnA[)], where NQ is the yield when the p-wave cross section op=0, n is the number of target nuclei/cm2, P„ is the neutron beam polarization, andA[ is the PV longitudinal asymmetry of eq. (1). For an unbroadened resonance, the relative difference of the neutron detector yields N+ and N. is (5)
£ =
N++-N__
_ , , / > „ . . _
>r , >r" N++N_
...
„/>,
=-tan(A[Pnnap)~-A[Pnnop,
where the Breit-Wignerp-wave resonance cross section is given by k
(E-Enf + (-f
Here r„ is the neutron width, T is the total width of the resonance, g is the statistical factor, k is the neutron wave vector, and Ep is the energy of the p-wave resonance. In practice, Doppler and instrumental resolution broadenings have to be included [3,17]. Correspondingly, for the P-odd T-odd D term we have AoPJ = ^
(7)
Im(/T -fo — lmD, k k where fr (/j.) is an amplitude for neutron spin parallel (antiparallel) to the direction of the vector 6 • (k x / ) . In a measurement, the P-odd P-odd observable is the quantity „PJ 0PJ+0PJ
P T
-Pj
i,pj
^{Es-Ep)
fj^
^
P T
where CT-J-' (CTJ,' ) is the total neutron-nucleus/7-wave cross section for a neutron spin parallel (antiparallel) to kxl, and v£,T is the nuclear weak matrix element of P-odd P-odd interactions between s- and p-v/ave resonances. According to eq. (1) and eq. (8), we can expect AP,T to be enhanced in p-wave resonances by the same factors as A1'.
167
The relative difference e ' of the neutron detector yields is
(9)
e ^ = fcA^_ A P ,r
Bunakov and Gudkov [11] have shown that (10)
AoPJ
yPJ =K—^AGP,
Vp
where K is a spin factor. Because the nuclear weak matrix element VF,T contains the same nuclear wave functions as V, we can define the quantity X as a ratio yP-r D (11) p V C ' The current experimental upper limits of the neutron EDM give a bound for the quantity X at the level of
2. An experimental setup to search for violation of timereversal invariance in neutron transmission A number of papers are proposing experimental arrangements to search for a P-odd T-odd violation in neutron transmission [14,15,19-22], However, no experiment has so far been done. The main difficulty has been to construct an experiment that would isolate the desired effect in the background of other much stronger interactions of the neutron spin, 6 • I, a B, and a • k. Other problems have been a lack of a high-intensity neutron beam and availability of polarized target materials that have/?-wave resonanances with large /"-violating asymmetries. Figure 1 shows a basic arrangement of a P-odd T-odd experiment. Small systematic errors in the experiment have proven to be difficult to control. These systematic errors may arise from: 1) deviations of the apparatus from rectilinear geometries, 2) differences in the analyzing power of polarizer and analyzer, 3) inhomogeneities in the target medium, 4) rotations of the neutron spin due to the holding field of a polarized target, (6-B), and 5) rotations of the neutron spin due to interactions of the neutron spin with the target spin, the Abragam rotation (
168
and detector, and rotation of the whole apparatus [19,20]. A true time-reversal operation corresponds to a reversal of the neutron momentum, the target polarization, and the neutron spin. In practice, the target polarization is flipped by an adiabatic rotation of magnetic fields by 180°. The neutron spin is flipped by reversing the polarization direction of the polarizer and analyzer using an adiabatic fast passage. The reversal of the neutron momentum is performed by rotating the apparatus 180° as shown in figure 1 [19,20]. This rotation also reduces the misalignment effects to a negligible level. Source
PT
Detector
"k,
a
Source
Figure 1. Basic arrangement of the f-odd T-odd experiment. CI and C2 are collimators, P is 'He polarizer, PT is a polarized 139La target, and A is 3He neutron spin analyzer. Reversal of the neutron momentum is carried out by rotating the apparatus 180°. The neutron spin and the target polarization are reversed by changing the direction of the currents in the magnetic field coils.
In the experiment, neutrons, emitted from an extended homogeneous source, pass through the collimator (CI), and are then longitudinally polarized by a 3He neutron spin filter (P). Next, the neutrons interact with a polarized target (PT) and
169 the holding field associated with the target. The sum of rotations due to the holding field and the Abragam rotation produce a rotation of approximately n as the neutron passes through the target. The acceptance of the neutron detector is defined by a second collimator (C2). Because the neutron spin rotates by n as it passes through the target, the neutron spin is approximately parallel or antiparallel to the direction of k x / at the midpoint of the target. The transmission of neutrons through the apparatus will be different around E=EP due to the absorption produced by the imaginary part of D. The detailed discussion of a sensitive search for the P-odd Todd coefficient D with this apparatus can be found in Ref. [19].
3.
About pulsed spallation neutron sources
The P-violation experiments and a search for a ^-violation in neutron resonances require high-intensity neutron fluxes to reach statistical goals. Since these experiments study neutron resonances, neutron energies have to be known accurately, which in spallation sources is possible through time-of-flight (TOF) measurements. Control of systematic errors also requires a good knowledge of neutron energies. The energy resolution in a TOF measurement depends upon the flight path length, the width of the proton pulse, the shape of the neutron pulse after the moderator, the electronics, and the detector. At low energies the main contributions to the instrument response function are from the intrinsic widths of neutron resonances and the moderator pulse shape. At high energies substantial time broadenings are introduced by the proton pulse width and the detector [24,17,3]. A typical epithermal neutron flux from a moderator of a spallation source has a Maxwell-Boltzmann distribution. In the epithermal energy region the flux, the number of neutrons with energies between E and E+dE per second on the detector, follows approximately a ME dependence and can be approximated by (assuming no material in the beam) (12)
^ , A T O ^
neutrons / Q
,
At E eV • s • sr where A'0 is a constant which depends on the power level of the source, AE is the energy width of a TOF channel,/is the fraction of the moderator surface viewed by the detector through the collimators, and Q is the solid angle of the transmission detector. Table 1 shows performance parameters of operating spallation sources, ISIS and LANSCE, and sources under construction, JSNS (Japan Spallation Neutron Source) and SNS. In table 1 the flux constants of eq. (12) is determined from the measured LANSCE flux and power [3] and then extrapolated to the power levels of the other spallation sources. In order to fully utilize the intense flux of JSNS and SNS, new transmission detectors have to be developed. The neutron rates on the detector will be so high that counting of neutrons does not work, instead a current mode detection has to be adapted. At LANSCE, with the 56kW source power, a 2-inch diameter scintillator at 56 m from the source had the instantaneous counting rate of 10 MHz [24].
170 Table 1. Performance parameters of spallation sources.
Source
Power (MW) ISIS 0.16(o) LANSCE 0.1(0) JSNS 1.0(d) 1.4(d) SNS (o)=operating power (d)=designed power 4.
Repetition rate (Hz) 50 20 25 60
Proton pulse width FWHM (ns) 400 125 600 700
V
(neutrons/s) 5xl0' 2 3xl012 3xl013 4xl013
Neutron beam polarizer and analyzer
The best way to produce polarized neutrons in a wide epithermal energy range is to transmit unpolarized neutrons through a sample of polarized protons [25] where the large spin-dependent n-p scattering cross section selects the neutron spin. Neutrons with their spin direction opposite to the proton polarization will be scattered and neutrons with the parallel spin direction will be attenuated by part of the triplet n-p cross section and scattering from other background nuclei present in the target. After the polarizer the neutron beam polarization is given by _ -sinhnAoP l + Pn° coshnAaPI cosh nAoP, - Pn° sinh nAoP, where Pi is the proton polarization, n is the number density of protons/cm2, Pn° is neutron polarization before the proton target, and Aa =(o+-o.)/2. Here o+ is a total cross section for neutrons polarized parallel and antiparallel to proton polarization. In the energy region from 1 eV to several keV, Aa has a nearly constant value of 16.7 b. With the thickness of lxlO 23 polarized protons/cm2 and with the proton polarization of 80%, close to 90% neutron polarization can be achieved. The protons in the frozen ammonia target were polarized by the dynamic nuclear polarization (DNP) method at IK and in a field of 5T [25,26]. Recently, a new type of DNP proton target has been developed. Protons in a crystal of naphthalene or p-terphenyl doped with pentacene have been polarized to 32% and 18%, respectively, at liquid nitrogen temperature and in a magnetic field of 0.3 T by means of microwave-induced optical nuclear polarization (MIONP) [27]. This target does not require a heavy cryogenic apparatus for the DNP process as does a conventional DNP target. In the DNP process paramagnetic centers, unpaired electrons, in the target material are used to transfer the electron polarization to the protons, but these paramagnetic centers also act as the primary source of the proton spin-lattice relaxation. In pentacene the paramagnetic centers are optically excited atoms. By turning off the optical pumping the source of the polarization
171 relaxation is removed and the proton polarization can be maintained in a low magnetic field. In a field of 7 gauss and at 77 K the polarization relaxation of 166 min in a naphthalene crystal has been reported [27]. A cell of polarized 3He is a nearly perfect neutron-spin filter or analyzer. The use of He to filter the neutron spin is based on the broad excited state of 4He compound nucleus with spin J"=0+. Neutrons with opposite spin to 3He will be absorbed and the neutrons with parallel spin will be transmitted through. The transmission of a polarized He cell can be expressed as (14) Tn = e""'T- (coshttAcrP, -Pn°sinhnAaPJ, where ore =(o++G_)/2. The transmission asymmetry of the cell is given by e = T+ ~ T~ = -Pn° tanh nAoP,. T++T_ The neutron beam polarization after the 3He cell can be calculated using eq. (13). The 3He spin filter has a number of advantages compared to the other few existing methods to polarize epithermal neutron beams. The He spin filter has: 1) a simplified experimental configuration, 2) compact experiments with polarizer and analyzer can be built [28], 3) the filter can be operated in a low magnetic field of 10 gauss or higher, 4) it has a large acceptance angle, 5) beam polarization can be measured very accurately with a transmission measurement [29], and 6) the neutron beam polarization can be flipped quickly with an adiabatic fast passage or adiabatic rotation of the He polarization. Because of the 1/v dependence of the neutron- He absorption cross section, the 3He spin filter is only practical below 10 eV. (15)
5.
Polarized nuclear targets for a triple-correlation experiment
The triple-correlation experiment requires a polarized nuclear target that has a neutron /j-wave resonance or resonances with known large P-violating effects. One of the best target candidates for the P-odd T-odd experiment is I39La. Its 0.748-eV p-wave resonance is the most studied of all those in which F-violation has been observed. The size of the longitudinal asymmetry is (9.56±0.35)xl0" [30,3]. A significant amount of effort has gone to the development of a polarized La target [31,32], Recently, 139La polarization of+47/-56% was achieved by the DNP method in LaA103 crystals doped with Nd3+ in magnetic field of 2.35 T and at temperatures below 0.3 K [33]. The operating mode of the 139La target in the P-odd T-odd experiment has to be a frozen spin mode where the DNP process, microwave pumping, will be turned off and the target will be cooled below 100 mK temperatures to reduce the spin-lattice relaxation time. Then the strength of the holding field can be decreased to a value where the spin-lattice relaxation rate is still small enough to allow the measurement to be performed. Available literature does not discuss relaxation rates of La in the Nd3+:LaA103 system at below 100 mK temperatures. The spin-lattice relaxation time constant of 82 min has been reported for La at 1.46K and in a field of 2.353
172 T [32]. This large relaxation rate is correlated to the strong electric quadrupole moment of 139La. m Xe also has been considered as a polarized target for the P-odd T-odd experiment because 13lXe has a 4% P-violating effect in the 3.2-eV p-wave resonance [34,35] and the ,29Xe isotope has been optically highly polarized [36]. Unfortunately, 13 Xe has a large nuclear quadrupole moment that prevents it from being polarized to a significant level [37]. 6. Summary An apparatus for studies of ^-violation in neutron resonances consists of a neutron spin filter, spin flipper, unpolarized nuclear target, and neutron detector. These components are well established also as the technique to extract the weak matrix elements from measured longitudinal asymmetries [3,7]. The intense SNS beam will take the P-violation measurements in neutron resonances to a new level of sensitivity where the experiments will be mostly limited by incomplete information available for some or all of the relevant spectroscopic quantities of the resonances. To date, no experiments have been performed to test the P-odd 7"-odd correlation in neutron transmission. The problems in building a sensitive experiment have been analyzed by several groups. A result from these discussions is an experimental design that diminishes false asymmetries caused by TRI neutron interactions coupled to misalignments of the components of the apparatus. The experiment will combine time-reversal transformations of the apparatus and rotations so that false asymmetries are reduced to the level where PPJ-violating signal can be detected. All the devices exist for this sensitive P-odd T-odd experiment, though some work still has to be done with the La target. The further development of the target should address the spin-lattice relaxation and feasibility to fabricate larger crystals suitable for the experiment. A lack of an intense epithermal beam has prevented earlier attempts to do a serious test of 77?/ in neutron transmission. The high-flux SNS and JSNS sources will remove this limitation. With the 1.4-MW SNS it will take about 10 days to reach the accuracy of 10"4 in X, assuming a 2-cm diameter and 4-cm long 139La target and 50% neutron polarization. It is fundamentally important to observe a direct violation of time-reversal invariance. However, we have to keep in mind that in analog with the P-violation studies in neutron resonances, we need several resonances that exhibit P-violation for obtaining statistically significant constraint on the strength of /"-violating interactions.
173 References 1. E. G. Adelberger, W. C. Haxton, Ann. Rev. Nucl. Part. Sci. 35 (1985) p. 501. 2. W. Haeberli and B. R. Holstein, in Symmetries and Fundamental Interactions in Nuclei, ed. W. C. Haxton and E. M. Henley (World Scientific, Singapore, 1995) p. 17. 3. G. E. Mitchell, J. D. Bowman, S. I. Penttila, and E. I. Sharapov, Phys. Reports 354 (2001) p. 157. 4. B. Desplanques, J. F. Donoghue, B. R. Holstein, Ann. Phys. 124 (1980) p. 449. 5. A. R. Berdoz et al., Phys. Rev. Lett. 87 (2001) p. 2301. 6. W. M. Snow, et al, nucl-ex/9704001, accepted for publication in Nucl. Instr. and Meth. (2002); W. M. Snow et al., in Fundamental Physics with Pulsed Neutron Beams, ed. C. R. Could, G. L. Greene, W. M. Snow (World Scientific, Singapore, 2001) p. 203; W. M. Snow et al, Nucl. Instr. and Meth. A 440 (2000) p. 729. 7. V. P. Alfimenkov et al, JETP Lett. 34 (1981) p. 295; V. P. Alfimenkov et al, Nucl. Phys. A398 (1983) p. 93. 8. G. E. Mitchell, in these proceedings. 9. P. Herczeg, in Parity and Time Reversal Violation in Compound Nuclear States and Related Topics, ed. N. Auerbach and J. D. Bowman (World Scientific, Singapore, 1996) p. 214. 10. D. E Groom et al. (Particle Data Group), Eur. Phys. J. C15 (2000) p. 1. 11. V. E. Bunakov, V. P. Gudkov, JETP Lett. 36 (1982) p. 32; Z Phys. A 3 0 8 (1982) p. 363. 12. P. Herczeg, Hyperfine Interactions 4 3 (1988) p. 77; P. Herczeg, in Tests of Time-Reversal Invariance in Neutron Physics, ed. N. R. Roberson, C. R. Gould, J. D.Bowman (Word Scientific, Singapore, 1987) p. 24. 13. V. P. Gudkov, Phys. Rep. 212 (1992) p. 77. 14. L. Stodolsky, Nucl Phys. B 1 9 7 (1982) p. 213; L. Stodolsky, Phys. Lett. B 172 (1986) p. 5. 15. P. K. Kabir, Phys. Rev. D 25 (1982) p. 2013. 16. J. E. Koster et al., Phys. Lett. B 2 6 7 (1991) p. 23; P. Huffman et al, in these proceedings. 17. V. W. Yuan, in these proceedings. 18. V. P. Gudkov, Phys. Lett. B 2 4 3 (1990) p. 319; V. P. Gudkov, Phys. Rev. C 46 (1992) p. 357. 19. J. D. Bowman, to be published. 20. Y. Masuda, Nucl. Phys. A629 (1998) p. 479c; Y. Masuda, Nucl. Instr. and Meth. A 402 (1998) p. 319; Y. Masuda, Nucl. Instr. and Meth. A 4 4 0 (2000) p. 632; Y. Masuda, in these proceedings.
174 21. A. P. Serebrov, JETP Lett. 58 (1993) p. 14; A. P. Serebrov, JETP Lett. 6 4 (1996) p. 1. 22. V. R. Skoy, Phys. Rev. D 5 3 (1996) p. 4070; V. R. Skoy, Nucl. Instr. and Meth. A 402 (1998) p. 323. 23. S. K. Lamoreaux, R. Colub, Phys. Rev. D 50 (1994) p. 5632. 24. Y.-F. Yen et al., Nucl. Instr. and Meth. A 447 (2000) p. 476. 25. S. I. Penttila et al., in Time Reversal Invariance and Parity Violation in Neutron Reactions, ed. C. R. Could, J. D. Bowman, Yu. P. Popov, (World Scientific, Singapore, 1994) p. 198; S. I. Penttila et al., in High Energy Spin Physics, ed. Kenneth J. Keller, Sandra L. Smith, (AIP Conf. Proc. 3 4 3 , AIP, New York, 1995) p. 532. 26. S. I. Penttila, in Proceedings of the 9th International Workshop on Polarized Sources and Targets, Indiana, 2001. 27. M. Iinuma et al, Phys. Rev. Lett. 84 (2000) p. 171. 28. T. Haseyama et al., Phys. Lett. B 534 (2002) p. 39. 29. D. R. Rich et al., Nucl. Instr. and Meth. A 481 (2002) p. 431. 30. C. D. Bowman, J. D. Bowman, V. W. Yuan, Phys. Rev. C 39 (1989) p.1721. 31. Y. Masuda, in Parity and Time Reversal Violation in Compound Nuclear States and Related Topics, ed. N. Auerbach and J. D. Bowman (World Scientific, Singapore, 1996) p. 83. 32. T. Maekawa et al, Nucl. Instr. and Meth. A 366 (1995) p. 115. 33. P. Hautle, M. Iinuma, Nucl. Instr. and Meth. A 440 (2000) p. 638. 34. J. J. Szymanski et al, Phys. Rev. C 53 (1996) p. R2576. 35. V. R. Skoy et al., Phys. Rev. C 53 (1996) p. R2573. 36. M. Gatzke et al, Phys. Rev. Lett. 70 (1993) p. 690. 37. R. Butscher, G. Wackerle, M. Mehring, J. Chem. Phys. 100 (1994) p. 6923.
T- VIOLATING THREEFOLD CORRELATION IN NEUTRON TRANSMISSION
Y. MASUDA Institute ofParticle and Nuclear Studies, KEK, 1-1 Oho, Tsukuba-shi, Ibaraki-ken, 305-0801 Japan E-mail: [email protected] A reciprocity theorem is applied to a T-violating neutron-transmission experiment. All the kinematical variables: neutron spin, momentum and nuclear spin are reversed in the application of the reciprocity theorem. The complete motion-reversal almost completely eliminates ^-invariant spurious effects. The methodology of nuclear polarization for the Tviolation experiment is also discussed. Many experimental attempts to find time reversal symmetry (7) violation have been made since the discovery of CP violation in the neutral Kaon system, which is the so far only known case of an observation of T violation (assuming the validity of the CPT theorem). T-invariance tests in neutron-nucleus forward scattering were proposed in Ref. [1]. A ratio of ^-violating amplitude to parity(/>)-violating amplitude, X is obtained in the tests. In the present, we discuss a possibility how one may detect the ratio X with the accuracy of hX < 10"4. The scattering amplitude of a spin 1/2 particle is a function of the neutron spin s and the nuclear spin / and can be represented in a general form in terms of the Pauli spin matrices cr„ ery and az and the identity matrix e: f=fo e + / ax +/ 2 cry +/ 3 az .
(1)
Particularly useful in order to see the T-violating contribution is the following decomposition of/ in terms of correlations between the kinematical variables k, s and /with definition of a coordinate system as shown in Fig.l [1]: As, k, I) =/o e +/,' *•/ +/ 2 ' s{kxl) +/,' *•* .
(2)
k is the neutron wave vector. The first and second terms in Eq.(2) are mainly due to the strong interaction. The nuclear polarization is kept by a magnetic field in the experiment. The interaction of the neutron spin with the magnetic field is included in the/!' term. The fourth term is P violating and induced by the weak interaction. Compared to the weak interaction between two nucleons, a very large enhancement by a factor of 106 is observed for neutron scattering at p-wave resonances on heavy nuclei [2,3,4]. The terms proportional to/,,/,' a n d / ' are invariant under motion
175
176 x A /
Fig.l Definition of a coordinate system for the description of the forward neutron scattering by a polarized nucleus.
^ * © — * /
Q
/
.
nucleus
Ax/
J'
reversal of all the kinematical variables k, s and / and therefore invariant under T. The third term in Eq.(2) changes sign under this operation and therefore is a Tviolating term. Many proposals have been made how to measure this contribution to the scattering amplitude [1,5,6,7,8,9]. A similar enhancement as for the Pviolating contribution is expected [10]. A description of methodological developments on nuclear polarization can be found in Ref. [11,12]. The effect of the/ 2 ' term on the neutron transmission should be observable as a transmission difference for the two inverse states of the neutron spin polarizer and the analyzer [1], whereby the sign of the spin correlation s-(Ax/) changes. However, in a real experiment different fake effects may occur. Usually the polarization of the neutron polarizer is different from that of the analyzer. This introduces a /-invariant /^'-coupling effect into the transmission difference, which is proportional to the difference between the polarizer and analyzer polarizations. A second fake effect is due to misalignments in the settings of the polarization axes of the polarizer, the analyzer and the target. Such misalignments induce an admixture of the amplitudes fi and f3J to the amplitude fj in Eq. (2) and may thus fake the /-violating observable in the transmission difference experiment. Similar fake effects are also found in the proposals of Ref. [7,8,9]. They impose serious constraints on the experimental technique such that it is very difficult to detect the /"-violating correlation if X is as small as 10"4 [13]. In the present article an extension of the transmission difference experiment [1] is given, which makes use of a reciprocity relation for the scattering amplitude (or the S-matrix), in order to suppress fake effects due to the misalignments. The reciprocity theorem, which will be discussed in the following, is based on the fundamental property of the inner product of two quantum states *FX and ^ [14]: = . <3) Neutron scattering can be described by the S-matrix which is represented by an inner product of an incoming wave tf/~ and an outgoing wave ¥\ which are taken as stationary states: < k'J, m'\s\ kj, rrO = < Vf,
j, .- I >T\
j;
„> .
(4)
177 \j, m > is a neutron and nuclear spin state. If the interaction is invariant under time reversal, the following expressions hold: TVtf ,
a
= exp(-
j_.a,
TY\
M
. = exp(-« >F _K , ; _ . ,
(5)
where >w is a phase relevant to the time reversal of the spin state. Using Eq.(3-5), a reciprocity relation is obtained for the S-matrix: < k'J, m'\s\ k,j, m> = exp((*V -fa < -k,j, -m I si -k\j, -trC > .
(6)
The scattering amplitude then fulfils the same relation: < k,j, nC l/l *,;, m > = exp(^. -^,) < -k,j, -m | / | -k,j, -rrC > .
(7)
The reciprocity relation for the scattering amplitude for a /"-invariant interaction, written in terms of the kinematical variables, is then given by As, k, I) =J{-s*, -k, - / * / =A-s, -k, -I) ,
(8)
where * and stand for complex conjugation and matrix transposition, respectively. The similar reciprocity relation as Eq. (8) is obtained for the S matrix. Considering the experimental apparatus shown in Fig.2, we can write the neutron transmission defining S-matrix of the whole apparatus as a product of S-matrices S, (i = 1, ..., n) which includes the effects of the polarizer, the analyzer and the target of polarized nuclei on the neutron transmission: S(k, / „ 72, ..., /„) = Sn(k, /„) - S2(k, I2) S,(*,h) .
(9)
The indices 1, 2, ..., n stand for the particles of all devices with which the neutron interacts. I, is the nuclear spin of the particle i. The 5-matrix is related with the scattering amplitude/as S,= \+2ikfik) ,
(10)
where / is the angular momentum of a partial wave./ has a similar form as Eq.(2), but now the kinematical variables are the neutron momentum and nuclear polarization only, since the incident neutrons are unpolarized. If we apply the reciprocity relation of the S matrix to Eq.(9), we obtain s(k, /,, i2 ,„ /„) = [S(-k, -/,*, - v , . . . , -/„*/]
178
S,(-k,-I,)S2(-k,-I2)
-
5n(-*,-/n)
(11)
Compared to Eq.(9), on the right hand side of the Eq.(l 1) the nuclear spins and the neutron momentum are reversed as well as the order of neutron transmission. It shall be shown below, that the fake effects due to the misalignments can be removed by application of the reciprocity theorem. Experimentally, the reversal of the nuclear polarizations can be realized by an adiabatic reversal of magnetic fields. The reversal of the neutron momentum and the order of neutron transmission can be realized by a rotation of the apparatus by 180° as shown in Fig.2 [15]. The rotated apparatus includes the collimators which define a neutron beam out of an extended and homogeneous neutron source. The detector is larger than this beam and assumed to have homogeneous detection efficiency. Fig.2 shows how neutron flight paths (indicated as arrows) in this beam behave under the rotation of the apparatus. The net result is an inversion of the neutron momentum. Misalignment in the rotation induces a deviation of the beam detector
Fig. 2 Neutron momentum reversal. The first picture is transformed into the second picture by a 180° rotation of the apparatus which includes the neutron collimators. The second picture is the same as the third picture because of rotation invariance. As a result, the rotation reverses the neutron momentum in the apparatus.
axis before and after rotation. However, the neutron count rate does not change if the neutron source and the detector are homogeneous. In the following it shall be argued that the effect of changed projection angles of the beam axis onto the neutron source and the detector due to the misalignment is a second order effect and therefore negligibly small.
179
The result is that all the ^-invariant misalignment-effects are cancelled by a complete motion reversal, i.e. inversion of all kinematical variables, which is performed by the reversal of the nuclear polarization and the rotation of the apparatus by 180°, according to the reciprocity relation. A simple calculation shows an experimental accuracy of 8/1 < 10"4 is possible. [16]. For the T violation experiment, the methodology of nuclear polarization has been developed at KEK. The 3He polarization, which uses a spin exchange with optically polarized rubidium atom, has been developed. Polarized 3He nuclei are applied to the neutron-spin polarizer and analyzer. A sapphire cell for 3He gas and a high-power diode-laser system are developed in order to obtain high neutron polarization. In Fig. 3, an experimental set-up for the 3He polarization is shown.
neutron spin neutron beam sappNireicel! Oth order diffraction 1800/mm grating
1 st order diffraction for laser cavity width < 0.15 nm
^
diode Saser width -1 nm
Fig. 3 3He neutron-spin polarizer and analyzer
A dynamic nuclear polarization for the lanthanum nuclear spin has been developed, since the largest /"-violating signal is expected in a p-wave resonance of 139 La.[17] In Fig. 4, typical NMR signals of lanthanum nuclear polarization in a LaA103 crystal are shown. In the left hand side, the signals of aluminum nuclear polarization are shown. The spin of 27A1 is 5/2. The six magnetic sub-states split in the quadrupole field of the crystal as shown in Fig. 5. As a result, five NMR lines are observed. The amplitudes of the NMR signals depend on the magnetic-substatepopulation distribution, which is explained by a spin temperature. The spin of 139La is 7/2, therefore, seven NMR lines should be observed. In the right hand side of
180 S?
AI
" 1 I1
I
I
I
25,90
26,00
I 14.10
14.00 frequency (MHz)
Fig. 4 NMR signal of polarized LaA10 3 crystal
*>AI SO* -2/5
**v
-3/2
'--.... >,^
-1/2
1/2 3/2 5/2
\ V
\
expi-En/kT)
~ - - „
**^
\ \
X, -1/4 fiaiyin1
population
magi letic sublevel
Fig. 5 Magnetic substate population
181 Fig. 4, one of the NMR signals is shown. Other signals are out of the scope. The nuclear polarizations were reversed by a suitable selection of microwave frequency. The result of the polarization reversal is shown in the lower side of Fig. 4. The lanthanum polarization was measured by a neutron transmission. The result is shown in Fig. 6. A transmission enhancement was observed at a 72-eV s-wave resonance, which corresponded to a lanthanum polarization of 10-20 %.
H
s wave resonance J=3. Eo=?2ev . I . • i . i • : • • i
, , I : , i , I
500
600
1.05
TOF channel
Fig. 6 Neutron transmission of polarized
La
A He- He-dilution refrigerator, which is shown in Fig. 7 is developed. The dilution refrigerator is essential to the T-violation experiment. In the ^-violation experiment, the neutron spin is vertical to the lanthanum spin. The neutron spin rotates around the nuclear spin in the presence of a pseudomagnetic field: real part of the nuclear-spin-dependent scattering amplitude. The rotation is cancelled with the Larmor precession, in order to keep the T-odd correlation in Eq. (2) during the neutron transmission. The magnetic field will be adjusted after switching off the microwave so that the pseudomagnetic rotation is cancelled. The temperature of the crystal should be kept low to hold the nuclear polarization. The 3He neutron polarizer and analyzer, and lanthanum polarizer will be mounted on a 3.5-meter-diameter rotating-frame as shown in Fig. 8. The neutroncollimation system will be also placed on the same frame. In the upstream an incident-neutron-monitor counter and in the downstream a neutron-transmission counter will be placed. 3He and 4He pump systems will be also placed away from the beam line.
182
Fig. 7 Dilution refrigerator for lanthanum nuclear polarization
Fig. 8 Experimental arrangement for the ^-violation experiment
183 The author would like to thank Prof. K. Morimoto, Dr. S. Ishimoto, Dr. S. Muto, Dr. T. Ino, Dr. J.D. Bowman, Dr. S. Penttila, Prof. G. Kim, and Dr. V. Skoy for their very useful discussions. Their comments and suggestions were indispensable for this work. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
L. Stodolsky, Nucl. Phys. B 197 (1982) 213; Phys. Lett. B 172 (1986) 5. V.P. Alfimenkov et al., Nucl. Phys. A 398 (1983) 93. Y. Masuda et al., Nucl. Phys. A 504 (1989) 269. CD. Bowman, J.D. Bowman and V.W. Yuan, Phys. Rev. C 39 (1989) 1721. P.K. Kabir, Phys. Rev. D 25 (1982) 25; Phys. Rev. Lett. 60 (1988) 686. J.D. Bowman, "Tests of Time Reversal Invariance in Neutron Physics" World Scientific (1987) 121. Y. Masuda et al., Hyp. Int. 74 (1992) 149; "Dark Matter in Cosmology, Clock and Tests of Fundamental Laws" Editions Frontiers (1995) 605. A.P. Serebrov, JETP Lett. 58 (1993) 14. V.R. Skoy, Phys. Rev. D 53 (1996) 4070. V.E. Bunakov and V.P. Gudkov, Z. Phys. A 308 (1982) 363. Y. Masuda et al., Proc. ICANS-XI, KEK Report 90-25 (1991) 1002; "Weak and Electromagnetic Interactions in Nuclei" World Scientific (1995) 58. Y. Masuda, Neutron Research 1 (1993) 53; H. Sato et al., Hyp. Int. 84 (1994) 205. S.K. Lamoreaux and R. Golub, Phys. Rev. D 50 (1994) 5632. M.L. Goldberger and K.M. Watson, "Collision Theory" John Willy & Sons (1964). The author is indebted to J.D. Bowman for the idea of apparatus rotation. Y. Masuda, Nucl. Instr. Meth. A440(2000)632. Y. Masuda, "Weak and Electromagnetic Interactions in Nuclei" World Scientific(1995) 58.
VIOLATION OF FUNDAMENTAL SYMMETRIES IN RESONANCE NEUTRON INDUCED FISSION A.BARABANOV Russian Research Center Kurchatov Institute,! Kurchatov square, 123182 Moscow, E-mail: [email protected]
Russia
W.FURMAN AND A.POPOV Joint Institute for Nuclear Research, 6 Joliot Curie street, 141970 E-mail: [email protected]. ru
Dubna,Russia
It is considered the multimode neutron induced fission going via neutron resonances (NR) as a unique possibility to observe and to study interference phenomena in nuclear fission or quantum mechanical aspects of the process. The P-even and P-odd angular correlations of fission fragments (FF) are discussed in terms of the formal theory of (n,f) reaction in a helicity representation developed recently by authors. Due to the proper account of total space parity in FF channels and specific advantages of the helicity representation for these exit channels it becomes possible to calculate the reduced fission amplitudes and to sum the general expression for the differential cross section of (n,f) - reaction over all non observed characteristics of FF. After the summation the observed angular correlations could be expressed via the reduced S-matrix defined with aid "integral" fission amplitudes specified for the J*K (A.Bohr's) fission channels. The conditions necessary for "survival" of FF angular correlations (after summation over huge number of exit channels with random fission amplitudes) are discussed in connection with a spontaneous violation of the shape symmetry of fissioning nucleus on the way from the normal to scission deformation. In all consideration it is assumed that space parity violation in fission process takes place at the stage of compound nucleus before an "exit" to the transition state of fission. Some examples of theoretical analysis of experimental data relevant to the subject under discussion are presented. The results show possibility to study a dependence of fission barriers on parity and to get more deep insight in the nature of transition states of fissioning nuclei. It means that such study provides some new tools to investigate properties of nuclei at extremely high deformation. In conclusion it is discussed some open problems as the tasks for future experimental and theoretical study.
1
Introduction
In spite of more than 60 years old history the quantum mechanical peculiarities of nuclear fission remain poorly studied. Any information that can be obtained on the properties of the wave function of fissioning nucleus which drives its transformation on the way from weakly deformed compound or ground state up to an extremely elongated one just before scission could be extracted only from a detailed study of characteristics of fission products and sometimes via prescission neutrons or gamma-transitions in second deformation well. (Last two possibilities are connected with serious experimental problems.)
184
185 The commonly used scenario of nuclear fission is next. On its way in "deformation space" a fissioning nucleus starting from the axial (A) and Rsymmetric (/?) (ellipsoidal) compound state penetrates or goes over the first fission barrier preserving the same shape symmetry. But after that the nucleus can change spontaneously its shape symmetry so in parallel with the former A and R symmetric state providing at scission so called symmetric fission modes it appears the additional states with only A symmetric (pear-like) shape leading up to asymmetric fission modes. This spontaneous violation of shape symmetry taking place at so called bifurcation points [1] belonging to many dimensional deformation space of fissioning nucleus brings to a very complicated structure of the two or three humped fission barriers. The location of these bifurcation points is under discussion up to now. Some experimental data [2] and its theoretical analysis [3] indicate that these points are situated in a vicinity of the second deformation well before the second fission barrier. If so the second (and third if it exists) potential well and the respective fission barrier(s) should be attributed to the definite fission modes. Taking into account that any axial symmetric states of deformed nucleus could be characterized by spin J and its projection K onto deformation axis as well as parity quantum number II (if it conserves) one can conclude that the structure of many dimensional fission barrier should depend on JTIK values. The wave function y3TmA (M is z-projection of J in lab reference system) describing transformation of fissioning nucleus on the way from a quasistationary compound state up to scission should be considered as a wave function of the transition state JTIKM defined into the whole configuration space from "an entrance" to the JTIK fission "channel" up to "an exit" for the fission product channel cf . This concept of multimode transition state [4] is a generalization of the A. Bohr's idea [5] of transition states defined originally at the saddle of (outer) fission barrier. Such transition states connected with the respective vibrational states [6] ^vib1™^ describe the large amplitude nuclear vibrations which lead a fissioning nucleus at the last end to its scission point. The vibrational states are coupled in turn with more complicated compound states attributed to the first and the second deformation wells. These compound states form successive stages of the hierarchy of various nuclear states involved into the fission process. ( For spontaneous fission one deals only with quasistationary states "living" in the first deformation well due simply to the fact that this well is deeper than the second one.) If one can study fission persuading via the transition states with a fixed or limited set of Jfl quantum numbers as it takes place for resonance neutron induced or spontaneous fission and in the case of well separated vibrational resonances there appears a chance to get some quantitative information on properties of the particular JTIK transition states. In such situation for (n,f)-process it is possible to observe [7,8,9,10] the interference of the reaction amplitudes corresponding to the same or different incident waves in an entrance neutron channel. These interference effects have very
186
complicated dependence on neutron energy which is rather different from the respective dependence of the total fission cross-sections. As it will be shown below the observed parity conserved (Peven) and parity violating (P0dd) interference effects are very sensitive to the shape symmetry of fissioning nucleus at scission and in general to the structure of the wave function of the transition states. A main goal of this paper is to discuss how the above mentioned hierarchy of the excited nuclear states related to the fission transition states could be investigated with aid of intensive pulsed neutron sources possessing various energy resolution. 2
Elements of formalism
The recently developed approach [11] to a description of nuclear fission induced by low energy neutrons based on the standard reaction theory joints naturally the extended concept of transition states [4] and a consistent description of all angular correlations of fission products observed in the experiments [7,8,9,10]. A general expression for the differential cross-section of (n,f)-reaction could be written in the form [11]
= ^2Z(g^)XZ*^fe^/»/fe/(0^;£j,
^ « " /
J',J
(i)
Ij'ljQ
where entrance channels are specified by neutron spin s, orbital / and total j momenta, by unit vectors of neutron momentum hk, by the same of target nucleus and neutron spin orientation, « 7 and ns respectively and by the incident neutron energy E„. The symbols gj denote the statistical weight of different total spins J of fissioning system. The unit vector n, describes relative position of fission products in an exit channel cj. A factor ^f,j.j'j{nknsnInfJ
incorporates all "geometric"
characteristics of the process. It has the form [11]
fo>W/)
®WSJ
= 4*r((2J +1)(27 + 1X2/ + l)(2y +1)(2/ + \){2s +1))^ x
Z ((2A + 1)(2A + 1X2JV + 1)(2M + \)/HNO {l)rM0 (s)C A0 * AhNM
J'
Jf
J Q
If
Jf
VI
j
I > <J
I
A
N
A
Ss
s A M
Z eSLPZuX* ("/ K . fa )YN„fa)YMW fa) qanXco
187
Here ZN0\I)
mdTMQ\S)
are spin-tensors of entrance channel and other symbols
are related with standard technique of angular momenta coupling. In formula (1) the term with Q=0 corresponds to the total fission cross-section and the following ones with Q^O describe an angular dependence of the respective differential crosssection. A reaction dynamics describes by the factor BJ [I'j'lj', En)
B ? ( / y , ; £ , ) = I i±H5B£ x n'n
^
J J
{i - 'cZolls;,(i'/En
-+FOC,II')SJ(!JEII
->FoC/n)+
F
(n'-n \K\>Q
' '~
F
which includes as key element S-matrix
Sj\ljEn
—> F KcAlj.
In the
framework of a modified helicity representation [11] the fission product channel c/ beside of internal quantum numbers of fission fragments is characterized by its total spin F, helicity K and total parity n . It is necessary to stress that if in course of (n,f)-reaction the parity n is not conserved and a respective elements of S-matrix are nonzero the above formulas describe simultaneously both the parity conserving and the parity violating processes. But in real experiments the characteristics of fission fragments cannot be fixed exclusively. Even for very elaborated fission product detectors a summation over large number of c, channels takes place. So to apply formula (1) to analysis of the measured total and differential fission cross-sections one has to sum the crosssection (1) over respective set of Cf channels. Really such summation should be done for the factor BQ (i'j'lj;En)
which depends on all characteristics of cf
channel including parities jtj and K2 of both fission fragments. It is impossible to obtain the formula for "observed" cross-section without introduction of some model for S-matrix. For resonance neutron induced fission the most appropriate model is the R-matrix one. In the frame of this model each cf channel is described by the real •TIF|£|<7
amplitude of a reduced fission width JV
attributed to the particular
188
compound level v. Using standard multilevel pole expansion of S-matrix we have in the last end for 2_,Be 'Q B
Q **l*r#
nf
where factors Pr
r*n'(Pfn'Pfn
r2'
2
(2)
>
are the R-matrix penetrabilities in cj channels. Due to the
fact that fission amplitudes JO ' '
are random for various compound states v
the sum in a right-hand side of formula (2) could be nonzero only for very special properties of these amplitudes. To clear out the problem let us calculate these values JTIFklc,
in the framework of R-matrix approach. By definition the amplitude JO is an overlap integral of the cj channel wave function and a respective function describing a fissioning nucleus at scission. The last function is the multimode wave function of transition state attributed to definite compound state v. In accordance with [12] and [4] this wave function could be written as =a™VfKU{{P)),
X™"
where the random factor a^ the wave function X „
(3) is an admixture of the transition state 4 0
to
of compound state v. If parity is violated in (n,f)-reaction
on the stage of compound nucleus so the wave function X v with both parities and instead of (3) we have
includes components
n Assuming that up to scission the transformation of fissioning system is adiabatic one we can use the collective model [13] for description of transition state wave function. So for the general case of only axial nuclear shape symmetry and K ^ 0 yrjJTlKM
_
2 ( ^ t l ) X {DJMK {0))Q>K
( A r)+n(-iy
+
*
DJM_K
(co)o^ (p, T)).
189
The internal wave function <E> K{j3,t) depending on internal variables r could be presented [4] as expansion over fission modes m
<^(/?,r) = £X(/0<MAr).
(4)
m
Here the functions OCm (/?) describe bifurcations of "trajectory " of fissioning system in course of its "movement" in multi dimensional deformation space {/?}. The wave function of fission product channel c/has the form:
^ | n -^
^ l
n
+ n ^ 2 (-1)" ^
),
(5)
0 f • ~. is defined as follows
where the function
JMF
\K\
KtK2
The wave functions of fission fragments XJK (7J [fly ]) depend not only on internal variables T/ but also on an orientation of the vector n,. Inserting the expression for
and (p f ,~,
X^
into the overlap integral defining an JUF\K\cr
amplitude of the reduced fission width JO
it is possible to obtain the next
result
Vl
h2n
Yl
./
rT -^ <^S^U
a-n )(!-*,*;)
2
•
<«
In this expression a^ and /if denote a channel radius and a reduced mass respectively. The factor tt^Li is a "weight" of the particular configuration of fission fragments c/-with total spin F and helicity l/sTl. It is essential to note that in course of calculation of the overlap integral we obtain that the following condition is fulfilled with good accuracy
K=K
190 which means that asymptotic quantum number helicity is fixed by projection K of total spin J of fissioning nucleus onto deformation axis. Now it is possible to finish a summation in formula (2)
Z
i ~i -r-i
jXl'lAK\cr'
Z, r*r
i ~i mF\K\cf
r*
i ~i nF\K\cf
p
f
(n'-n)i,» 2 .
l
;
n'-n .—J~
=l
.m'\K\ JP,n\K\
y*m Y™ ' <7>
Cj-czm F
where the next definition for "observed" reduced width is introduced:
rf^^^lSZ^^,) 2 ^^ cfczm F
.
(8)
^-Mf
The results (7) allows one to preserve after summation over non observed fission channels cj all interference terms in the differential cross-section of (n,f)reaction (1) in the multilevel many channel approximation. But instead of S-matrix defined in a space of fission product channels cj we have to use the reduced Smatrix defined in the space of JITK channels. This result provides a consistent justification of the A. Bohr's hypothesis [5] of transition states and the Reich-Moore approximation [14] for reduction of S-matrix. It is necessary to note that a "survival" of P even and P odd interference effects demonstrated above becomes possible due to a specific parity dependence of the transition state wave function
*F^
. Namely the wave function 0 ^ ( ^ , 7 )
of
fissioning nucleus as defined in its internal coordinate system does not depend on total parity n . It describes pear-like shape of nucleus at scission and provides the asymmetric fission modes. But in the case of the ellipsoidal nuclear shape preserving J? symmetry it is not the fact. So factors 0Cm and uJK< become parity dependent and sums in (2) and (7) goes to zero. This case usually attributed to symmetric fission modes. But R symmetric scission configurations should be connected with only "true" symmetric fission when both fragment are identical. Really the symmetric fission includes mostly close in charge and in mass fragments with different other characteristics. So a prescission nuclear configuration should have the wave function similar to the case of asymmetric fission. And all interference effects should be observed.
Discussion With aid the formalism considered above it is possible to derive the formula for analysis of various interference effects discussed in the section 1. This is
191
+ < ) f l L (En )p„ (nf [nkns ]) +
(9)
where
< = * x2E^ZEl SAVE,, -> W)|2= X
n
/>
A-
j
n
is the total fission cross section expressed as a sum of the spin-separated components. The factor p„ is neutron beam polarization and f2 is an alignment of target nuclei. The terms for forward-backward (FB) and right-left (RL) correlations could be easily obtained by a straightforward calculation. Above formulas were used for analysis of experimental data on angular anisotropy of fission fragments from resonance neutron induced fission of aligned nuclei 235U [15] and on P-even angular correlations for 235U(n,f)- reaction [10]. These data was fitted simultaneously to reproduce the total fission cross-section as well as the existing data on neutron transmission and radiative capture. In Fig. 1 an example of such fit is presented. A satisfactory fit of the data was achieved. But some uncertainties in the obtained parameters of p-wave resonances remain. There are too many free parameters involved into the fit so even position and total number of these resonances are something indefinite. But average parameters of p-wave resonances obtained in such a way in comparison with respective parameters for swave resonances could give more reliable information on the parity dependence of fission barriers for JT1K channels. In general due to very irregular energy dependence of the coefficients of Peven and Podd angular correlations of fission fragments from resonance neutron induced fission a study of this effects provides a sensitive tool for investigation of fissioning nuclei. Up to now there was made only first steps on this way. To obtain more definite and quantitative information it is necessary to continue these investigations with more intensive neutron beams and with better energy resolution. Detailed study of neutron energy dependence of mass&TKE distributions of fission products as well
192
Figure 1. The fit of experimental data on the a™ (E,,) and aLR (E„ ) dependencies for 235U(n,f)-reaction with account of the p-resonances shown in lower part of the picture. The points are experimental data [10] and the curves are results of present fit.
as fragment mass dependence of different angular correlations for wide interval of incident neutron energy could permit to understand deeper many open problems. Among them it should be mentioned - the nature of bifurcations of trajectories in deformation space of fissioning nucleus; a location of bifurcation points for different fission modes - the multimode structure of second and third fission barriers; their dependence on value of K projection and parity IT. as well as on other quantum numbers - the angular correlations of fission fragments in the resonances of second deformation well -the fluctuations of mass&TKE distributions of fission products for statistically meaningful set of neutron resonances - the properties of vibrational resonances in neutron induced fission and its interconnections with fission modes All these problems could be studied with aid of new powerful pulsed neutron sources such as SNS at ORNL , USA and at KEK, Japan. Some particular investigations could be realized at existing neutron sources - LANCE, LANL,CERN n-TOF facility, ORELA, ORNL and also at IREN source, JINR, Dubna being under construction now.
193 Acknowledgements The work was supported by the grant of INTAS #99-0229. References 1. Brosa U., Grossmann S. and Mueller A., Phys. Rep., C197 (1990) p.167 2. Hambsch F.-J., Knitter H.-H., Budtz-Jorgensen C. and Theobald H., Fission mode fluctuations Nucl. Phys. A491 (1989) pp. 56-90 3. Furman W.I., Fission channels and modes. In Proceedings 1999 Frederic Joliot - Otto Hanh Spring Session on: Neutron Data Measurements andEvaluation, (Institute for Reference Materials&Measurements, Geel, Belgium, 1999) pp. 125-145 4. Furman W.I. and Kliman J., Fluctuations of fission characteristics and the structure of fission channels. In Proceedings of 17th lnternatinal Symposium on Nuclear Physics, ed. by D. Seeliger and H. Kalka (ZfK,Dresden, 1988) pp. 142-147. 5. Bohr A., On the theory of nuclear fission. In Proceedings of International conference on the Peaceful Uses of Atomic Energy, (United Nations, New York, 1956) v.2, pp. 151-153 6. Bjornholm S. and Lynn J. E., Rev. Mod. Phys. 52 (1980) p.725 7. Pattenden N.J. and Postma H., Nucl. Phys. A167 (1971) p.225 8. Tambovtsev D.I, Kozlovsky L.K., Gonin N.N. et al., Phys. At. Nucl. 60 (1997) p. 877 9. Danilian G.V., Vodennikov B. D , Dronyaev V.P. et al. Pis'ma ZETPh 54 (1977) p. 9 10. Alfimenkov V.P, Chernikov A.N, Lason L. et al, Nucl.Phys., A645 (1999) p.31 11. Barabanov A.L. and Furman W.I. Formal theory of neutron induced fission. Z ^/zys., A357(1997) pp. 411-418 12. Porter C.E. and Tomas R.G, Fluctuations of nuclear reaction widths. Phys. Rev., 104 (1956) pp. 483-491 13. Bohr A. and Mottelson B.R, Nuclear structure. (W.A. Benjamin, Inc., New York, Amsterdam, 1974) v.2 14. Reich C.W. and Moore M. S, Multilevel formula for the fission process, Phys. Rev. I l l (1958) pp. 929-933 15. Kopach Yu.N, Popov A.B, Furman W.I. et al. The investigation of the fragment angular anisotropy in resonance neutron induced fission of the aligned 235U target and a role of JK channels, Phys. At. Nucl. 62 (1999) pp. 900915
PHYSICS OF THE FISSION PROCESS A N D P A R I T Y VIOLATION IN N E U T R O N INDUCED REACTIONS
VLADIMIR GUDKOV Department of Physics and Astronomy University of South Carolina Columbia, SC 29208 E-mail: [email protected]
1
Introduction
Parity violation in nuclear fission has been observed 1 about 25 years ago. This is a rather long period in t h e whole 64 years of history of nuclear fission since its discovery in 1938. Therefore, one can ask what we have learned new about fission from parity violating effects and what could be done with the new generation neutron sources. T h e nuclear fission process is widely used in modern technologies. However, theoretical description of t h e fission is far from complete despite continuous study of it for many decades. One of t h e reasons for t h a t is a very high complexity of t h e nuclear fission process, which includes all difficulties of the nuclear theory from nuclear forces descriptions u p to the theory of q u a n t u m many body systems with collective motion modes. Simple models 2 ' 3 ' 4 , which where proposed many years ago, can describe many key features of the fission process. These models were enough for many technical applications, provided the experimental tests of calculations and simulations were available. Recent progress in t h e development of high power computers and more sophisticated applications of nuclear physics make numerical simulations of nuclear process more important for nuclear systems designs. W i t h the demand for better accuracy and reliability of the simulation codes, t h e simple models of nuclear fission are insufficient in many cases, for example, in calculation of angular and mass distributions of t h e fission fragments. From this point of view the study of parity violating effects in nuclear fission looks very promising since these effects, being of tiny differences in t h e angular distributions (asymmetries) of fission fragments with different masses, are very sensitive to t h e nuclear fission models. It is well known t h a t t h e investigation of t h e parity violation in heavy nuclei is not t h e method to study properties of weak interactions but rather
194
195 it could be a powerful tool t o study nuclear structure and nuclear reaction mechanisms. Parity violation in neutron induced fission gives additional opportunities t o study a process of transition from complex compound (totally statistical) nuclear states to a strongly correlated q u a n t u m motion of deformed nuclei, which decay into fission fragments. This transition, usually called a fission mode, was t h e source of m a n y surprises in t h e nuclear fission process. For example, it was expected t h a t t h e distribution law of the resonance fission widths would be similar to t h e widths distribution law for 7-decay in neutron induced (n, 7) reactions or, even more, would b e practically a 5-function type distribution. This is because the number of final states of fission fragments Nf ~ 10 1 0 (taking into account different masses, charges, excitation energies, etc.) is much larger t h a n t h e number of final states for (71,7) reaction. T h e widths distribution is a %2 distribution with t h e number of degrees of freedom equal to the number of decayed channels, and fluctuations of widths in this distribution decrease with increasing t h e number of channels. However, the experimental fission widths distributions look like t h e neutron widths distributions which are t h e Porter-Thomas distribution, or the x 2 distribution t h e with one degree of freedom. This shows t h a t the 'effective' number of degrees of freedom for fission should be very small. Another puzzle to mention is the unusual interference in t h e total (integrated over the angles) fission cross sections among compound resonances with t h e same spins. To explain these unusual effects, t h e Bohr's hypothesis had been suggested (see, for details refs. 2 ' 3 , 4 ' 5 ). According t o this hypothesis, t h e fission process could be described in terms of a few fission transition states, which play a role of degrees of freedom for the fission and have the fixed set of q u a n t u m numbers: t h e compound nuclear spin, its projection on t h e nuclear symmetry axis, and parity. T h e similar approach 6 based on the existence of t h e fission transition states helped solve one more puzzle with the anisotropy of t h e angular distributions in 7-fission and fast neutron fission. T h e discovery of large parity violating asymmetries in resonance fission made it necessary to add new features to the 'conventional' fission model 7 and rose u p new questions about t h e Bohr's hypothesis 8 in order t o explain non-vanishing effects in the interference of channels with opposite parities. We will not go deep into the details of theoretical models for the description of parity violation in fission (for detailed information see, for example 7 , 8 ' 9 and references therein, and the recent approach presented by W . F u r m a n at this Workshop 1 0 ) but rather consider some general features of the process using a schematic description. Before we discuss t h e parity violation in neutron induced reactions, let us mention the possibility of parity violation in spontaneous fission. P-odd cor-
196 relations in angular distributions of fission fragments have been predicted 11 a long time ago with the assumption that weak interactions mix wave functions of opposite parities of the deformed nucleus. The pear-like shape of nuclear deformation, with the axis being coincided to the direction of fission fragments momenta pf, oriented along the spin I of the nucleus leads to the P-odd correlation (I -pf) in the angular distribution of the fragments. Then the P-odd correlation is proportional 11 ' 12 to the ratio of the weak matrix element between the deformed states with opposite parities and the distance between these states. Also, it is proportional to the ration of barrier penetrabilities for the state with opposite parities. The barrier penetrabilities play the main role in the enhancement of the P-odd effects in spontaneous fission since their ratio could be as much as a several orders of magnitude due to the difference in penetrabilities for states with opposite parities. This factor could be important for a neutron induced sub-barrier fission if one state (or both states) is under the fission barrier. Also, it could be an additional barrier resonance enhancement in the case of a double structure of the fission barrier 12 . Despite the prediction of P-odd effects for spontaneous fission it was unexpected that the P-odd effects in neutron induced reactions would be on the level ~ 10~ 4 because the number of final states for fission Nf is very large and, therefore, the simple statistical averaging over the final states would make any manifestation of parity violation vanish. This is, again, the point where the coherence of the process (like the Bohr's hypothesis) has to be applied to explain the large experimental effects. 2
Angular Correlations in Fission
Let us recall some features of parity violating effects in neutron induced fission using as an example a P-odd correlation between neutron spin 5 and the light fragment momentum pf. anjiS ~ ( • Pf). Assuming that parity is mixing on the compound nuclear stage and taking into account only two nearest resonances with opposite parities (s- and p-wave neutron resonances) one can obtain 8 ^ a
n,fis
^
e
l^LfXJssfps
r
+ fppfsp}
,. .
,
(1)
where ajiS is a total fission cross section, fss and fpp are parity-conserving amplitudes, / p s and fsp are parity-violating amplitudes:
'- ~
'r'''/V"/'(g^,1+,r,/2)'r"'1/Va:i-
<2>
197
P
W » p
s
/ (n \
\ W f) t / s p
Figure 1. Parity conserving fSs (diagram a) and / p p (diagram b) and parity violating fps (diagram c) and fsp (diagram d) partial fission amplitudes.
'-~
f U
(r?)'^» ( E _ E ; + i r p / 2 ) ( r;)-/v^,
^ (rf\l/2(i5{) {ll)
ZK (rn)l/2(i5;) e P (E-Es + irs/2)(E-Ep + iTp/2){1^ -
(3)
(r\ (5)
Here TJ is the partial width of the compound resonance of the decay into fragments in state / with relative angular momentum I, and df is the potential phase shift in this channel; I' = I + 1. r™p and 5" are neutron width and potential phase for the s- and p-wave compound resonances with the resonance energies ESjP and the total widths TSiP. E is neutron energy and W is parity violating mixing matrix element for s and p compound resonances. These amplitudes can be represented by diagrams on Figure 1, where symbols s and p correspond to s- and p-wave propagators 1/(E — ESiP + iTStP), vertexes correspond to decay amplitudes (rf) 1/,2 exp(<5f) for a channel k with parity i, and W is the parity mixing matrix element. From eqs.(l)-(5) one can see that for low energy neutrons the first term in eq.(l) is dominant since it is proportional to T™ (fssfps ~ F") but the second one is proportional to T" (fppfsp ~ F™)), which is smaller by orders of magnitude.
198
Another observation from these expressions is that the parity violating part is separated (factorized) from the fission process due to the parity mixing on the compound nuclear stage (it could not be the case for spontaneous or sub-barrier fission). Prom this point of view the mechanism of production of P-odd asymmetries, which is caused by the interference of fission decay channels with opposite parities (relative orbital momenta of fragments) should be the same as of the P-even symmetries. For example, for the left-right (P-even) correlation (a • [p"n x p)}) (where p"n is the neutron momentum):
aLR ~
^m^2lJ{fsj;p}/(2afis).
The sum over all states of the fission fragments in the eqs.(l) results in the term ~ X ) M ( / / } ( M I;, ) 1,/2 e^ i'~ ' ', which vanishes if the fission widths of the fragments with opposite parities are uncorrelated. Therefore, one needs to introduce new correlations for the fission amplitudes to explain the observed non-zero asymmetries. It could be a generalization of the Bohr's hypothesis in terms of sign correlations of fragment's amplitudes 8 , or a model of fission involving a pear-like deformed transition states 7 , or a coherent summation over fragment's amplitudes in the helicity representation 10 . The common feature of these approaches is a very strong coherence of the process of the fission as opposed to almost pure statistical behavior of nuclei on the compound nuclear stage. As a result of a transition of a quantum system from a random statistical state into the 'self-organized' highly coherent system, one gets £jJ(j,)( r f r £) 1 / 2 e < ( * , / '~*' / ) - ( r ^ r ^ 8 ) 1 / 2 , where r £ £ is the total fission amplitude. Then, taking into account only the product of two amplitudes fss and fps (diagrams a and c) in eq.(l), one obtains 8 anJlsC
2W(E - Ep)
"(E-Ep)*
(T^_
+ ri/4{rfjs
1/ *
(6)
One can see that due to the mixing on the compound nuclear stage, the P-odd effects are enhanced in the vicinity of p-wave resonance. However, due to high resonance densities and large total widths for fission nuclei (for example, 2 3 5 [ / and 2 3 9 Pu), the enhancement factor is numerically smaller than the resonance enhancement of P-odd effects in neutron transmission. It should be noted that values of P-odd and P-even asymmetries have the same order of magnitude in the eV neutron energy region. This is mostly because of a coincidence in the numerical values of neutron barrier penetration factors for P-even asymmetries and the parity mixing parameter for P-odd asymmetries for the heavy fission nuclei (for details see, for example references 8 ' 13 ). Therefore, one can see that both P-odd and P-even asymmetries could provide similar information about the fission process. Moreover,
199
sometimes P-even asymmetries could be even more preferable than P-odd ones because they increase with the neutron energy faster, do not contain unknown weak matrix element, and, what's more, may not require polarized neutrons (for example, (p~n • p}) correlation). However, the similarities between P-odd and P-even effects (from the point of view of the fission process) are not so straightforward and simple even in the two-level approximation. One example would be the case of very small j9-wave neutron resonance (r£ ~ 0), where the P-even asymmetries became zeros but P-odd anjiS is not affected by the value of T£ at all. Parity mixing on the compound nuclear stage results in the very strong energy dependance of the P-odd effects on the scale of the average distance D between neutron resonances. The excitation energy of compound nuclei has a scale of about ~ 6 MeV. Therefore, a small change in neutron energy (.D ~ 1 eV) could lead to an essential change of the parity mixture but it does not change essential nuclear parameters. For the over-barrier fission, this will affect only the value of P-odd effects due to resonance enhancement 8 . However, for the case of a sub-barrier fission, the process would be more complicated due to barrier penetration factors. In such cases, the measurement of the energy dependencies of P-odd and P-even effects could be a sensitive tool to clarify details of the fission process and new possible mechanisms of the parity mixing. Up to this point, we discussed asymmetries in fission with a crude separation of fragments into two groups: the light and the heavy ones. To understand the fission process one must be able to describe not only integrated features but also more detailed properties. The next step would be a study of double cross sections such as a dependance of P-odd and P-even asymmetries on masses or charges of the fission fragments. These parameters are very important since they can clarify the mechanism of the fragment formation. For example, statistical models assume the formation of mass distributions at the scission point but angular distributions are formed before it. Therefore, for that class of models one cannot expect differences in asymmetries for fragments with different masses. Some measurements (see, for example 14 ' 15 ' 16 ) of these mass dependencies have been done. However, more data and better accuracy are needed for an unambiguous analysis of the fission models. Another process to study the fragment formation mechanism is a ternary fission. The branch ratio for the ternary fission is usually less than 1%, however, there are more angular correlations available because the third fragment (usually Q-particle) brings additional degrees of freedom. For example, it is possible to have P-odd correlation even with unpolarized neutrons 9 : Pa " [Pn x Pf]i where pa is momentum of a-particle.
200
This correlation is P-odd and T-odd one (in the sense, that it changes the sign with the changing sign of the time variable). However, it does not related to time reversal invariance (TRI). This statement is trivial but should be addressed here, since in the last years some publications (see, for example refs. 17 ' 18 ) seriously discuss possibilities to test TPJ by measuring T-odd correlations in ternary fission. They consider the P-even T-odd correlation a • \pf x pa]. One should say that the statement that this (or other T-odd) correlation could be related to TRI violation is absolutely wrong and misleading. The simple way to show this (see, for example a general theory in the classical quantum mechanics textbook 19 and its applications to neutron induced reactions in ref.20) is to recall that parity invariance and TRI are essentially different. TRI leads to no restrictions on the reaction amplitude, but rather to a relation between amplitudes for different processes. Therefore, it does not forbid T-odd terms's presence in the amplitudes and in observed parameters at all. However, for a special process which could be described almost completely in the first Born approximation, it is possible to relate time reversal non-invariant parameters to T-odd terms in the amplitude of the process. Since it is impossible to describe the fission process in the first order of the Born approximation due to the high level of the process complexity, the relation between T-odd correlations and TRI cannot exist for the fission process under any circumstances. Theoretical approach 9 to the estimation of P-odd effects in the ternary fission has been suggested many years ago. Under reasonable assumptions about the fission process, P-odd effects in ternary fission were predicted to be on the same order of magnitudes as the effects of the binary fission (the possible suppression factor for the P-odd effects in ternary fission, in comparison to the binary one, would be an indication of different coherence processes for these two fission modes). Also, it has been shown that P-odd effects in ternary fission have similar resonance enhancement factor as in the case of binary fission. Current experimental measurements of P-odd effects in ternary fission21,22 could be considered a confirmation of the similarity between two fission modes, however, more precise data are needed to make definite conclusions. 3
W h y we need the SNS to study asymmetries in fission?
After this short review of the current situation with the study of P-odd (and P-even) correlations in fission, one can see that these correlations are very sensitive to the fission mechanisms. Therefore, one needs more data for different targets with over-barrier and sub-barrier fission, and for binary and
201 ternary fission. Also, one needs to measure not only 'averaged' asymmetries but rather double differential fission cross sections with high accuracy and in a wide energy range. The high flux neutron beam with a reasonable energy resolution is the key factor for these future experiments. It is very important to increase the current accuracy, at least by an order of magnitude, to be able to answer important questions related to the fission dynamics. The obtained information would be of great interest for the nuclear theory, and it will lead to improvements of nuclear fission models, which are required for simulation codes used in many nuclear applications. It would also contribute to our general understanding of quantum theory of transitions from pure statistical to highly coherent states. References 1. G.V. Danilyan et al, JETP Pisma 26, 298 (1977). 2. A. Bohr in Proc. Int. Conf. on peaceful uses of atomic energy, vol. 4, p.220, (Geneva, 1955). 3. J.E. Lynn, The Theory of Neutron Resonance Reactions, (Clarendon Press, Oxford, 1968). 4. A. Bohr and B.R. Mottelson Nuclear Structure, (W.A. Benjamin, New York, 1969). 5. V.L. Sailor, see ref.[2] p.192. 6. V.M. Strutinsky , JETP 39, 781 (1960). 7. O.P. Sushkov and V.V. Flambaum, Usp. Fiz. Nauk. 136, 3 (1982). 8. V.E. Bunakov and V.P. Gudkov, Nucl. Phys. A403, 93 (1983). 9. V.E. Bunakov and V.P. Gudkov, Z. Phys. A321, 271 (1985). 10. A. Barabanov, W. Furman and A. Popov in This Proceedings. 11. V.V. Vladimirsky and A.N. Andreev, JETP 41, 663 (1961). 12. A.P. Budnik and N.S. Rabotnov, Phys. Lett. B 46, 155 (1973). 13. V.P. Gudkov, KEK Report 91-2, (KEK, Japan, 1991). 14. A.K. Petukhov et al, JETP Pisma 30, 470 (1979). 15. A.Ya. Alexandrovich et al, Nucl. Phys. A576, 541 (1994). 16. A. Kotzle et al, Nucl. Instrum. Methods A440, 750 (2000). 17. G.V. Danilyan , JETP Lett. 70, 565 (1999). 18. P. Jesinger et al, Nucl. Instrum. Methods A440, 618 (2000). 19. L.D.Landau and E.M. Lishitz Quantum mechanics: non-relativistic theory, (Oxford, New York, Pergamon Press, 1965). 20. V.P. Gudkov, Phys. Rep. 212, 77 (1992). 21. A.B. Belozerov et al, JETP Pisma 51, 10 (1990);54, 132 (1991). 22. F. Gonnenweina et al, Nucl. Phys. A576, 303 (1994).
POSSIBILITIES FOR STUDIES OF PARITY VIOLATION AT THE SNS USING THE CAPTURE GAMMA REACTION A.C. HAYES Theoretical Division Los Alamos, NM 87545, E-mail: [email protected]
USA
LUCA ZANINI CERN, CH-121I Geneva 23 Switzerland E-mail: [email protected] We have shown that in the standard resonance theory of parity violation in the compound nucleus, the longitudinal parity violating asymmetry in partial neutron capture cross sections is essentially independent of the partial gamma-widths involved. Thus, the same asymmetry is expected for each partial cross section involving the same p-wave resonance. The asymmetries are expected to be enhanced (~ 10%), and asymmetry measurements for several partial capture cross sections from a given p-wave resonance would provide a very strong test of the theory. Measurements of parity violation in the (n,y) reaction using high efficiency germanium detectors at the SNS could determine the parity-odd nucleon-nucleon matrix elements in complex nuclei with high accuracy and shed light on the origin of the sign problem observed in 232Th. Additionally, simultaneous studies of the El and VPNc matrix elements involved in these decays could be used to help constrain the statistical theory of parity non-conservation in compound nuclei. We propose a class of experiments and examine their feasibility.
1
Introduction
In neutron transmission experiments on heavy nuclei, the parity-violating asymmetries, which are defined as the fractional difference of the resonance cross section for neutrons polarized parallel and anti-parallel to their momentum, An
0+-<5~
c +c can be as large as 10%. These represent by far the largest parity violating asymmetries observed in nuclei. The measurements have been carried out by the TRIPLE collaboration [1] on p-wave resonances in compound nuclear systems such as 238U, 232Th, 108-106pd, and 93Nb. In these systems the energy separation between opposite parity s-wave and p-wave resonances ranges from 0.1- 100 eV. The parity violating mixing of s-wave states into a /?-wave state leads to a longitudinal asymmetry
A.=2yfelE
202
(2)
203
Here T" and T" are the neutron partial width amplitudes for the p- and s-wave resonances, and \0 S |V PNC |(|) y is the matrix element of the two-body PNC NN interaction between these resonances. The large size of the PNC asymmetries in compound systems arises in part because of the small energy denominators involved and because of the very favorable ratio of s-wave to p-wave neutron widths. One of the main advantages of the TRIPLE program was the ability to measure PNC asymmetries on several resonances in the same nucleus, thus allowing a likelihood analysis of the data to extract the mean squared PNC matrix element, M = | V PNC |
.
The
mean
squared
matrix
element
is
defined
as
2
M = (1 / N p N s ) £ s p u t ) s | VPNC|{f)p V where Np, Ns are the number of/?- and .?wave resonances occurring in a specified energy window. Here we examine the possibility of a complementary systematic study of PNC in compound nuclear resonances using the («,y) reaction. In particular we emphasize the advantages of partial capture cross section measurements, where PNC asymmetries could be measured for several gamma-decay branches of the same p-wave resonance. 2
Parity Violation in Neutron Capture
The general expressions for parity-odd (P-odd) correlations in the («,y) reaction have been derived by Flambaum and Sushkov (FS)[2]. We concentrate on the Podd correlation in the (w,y) cross section that depends on the neutron helicity and direction of the neutron's momentum, namely, a.k n . The dominant contribution to parity violation in neutron capture is usually assumed to arise from mixing between the p- and s-wave resonances in the entrance channel. There are two P-odd amplitudes that contribute to the asymmetry. In the first (denoted V3 by FS) the neutron is captured by an s-wave resonance and the photon is emitted by a P-odd pwave component in the resonance wave function. In the second (denoted V4) the neutron is captured by a p-wave resonance and the photon emitted by a P-odd swave component. The amplitudes for both of these involves a product of matrix elements for a strong interaction neutron capture, a weak interaction parity mixing, and the photon emitted by a P-odd s-wave component. The amplitudes for both of these involves a product of matrix elements for a strong interaction neutron capture, a weak interaction parity mixing, and electromagnetic interaction photon emission. The full expressions for V3 and V4, and the corresponding P-even amplitudes for s(jE?)-wave neutron capture with emission of the photon from the same s(p)resonance (V,(V2)), are given in FS. The a.k„ P-odd correlation in the cross section is then 2Re(V2V3' + VjV4*). Starting with these four amplitudes, we obtain an expression for the helicity asymmetry ALV in partial neutron capture,
204
A? (3)
Here E is the neutron energy, r n p , r n s are the neutron partial widths, and Tp, Ts are the total resonance widths. The index /' appearing in (3) refers to i'h gamma transition from the p-v/ave resonance under consideration. The partial gamma widths Vp are for individual gamma transitions from the same p-wave resonance to different final states. Expression (3) is completely symmetric in p and s and is exactly analogous to the corresponding expression for neutron transmission, (see eq (28) in Bunakov and Gudkov [3]). If F^p/F^s >1 a good approximation to Aj1 is obtained by setting E=EP and neglecting the total widths, Ts and Tp, in the denominator, in which case the longitudinal asymmetry in neutron capture becomes
fr 7 ^
= 2
^
(Es-EP)
The average ratio of the El strength from p-wave resonances to the Ml strength from 5-wave resonances for primary gamma-rays is typically greater than one. In such cases eq. (4) is a good approximation, and AL7 is independent of the partial y-widths involved. The longitudinal asymmetry corresponding to the observable o.kn takes on the same value in both partial neutron capture and transmission measurements, and is quite enhanced in both cases. The TRIPLE collaboration usually quote cross section asymmetries as a fraction of the parity conserving pwave cross section op(E), as opposed to as a fraction of the total cross section at Ep. In this case the expression for A7L is simpler that eq. (3) and is not symmetric in/? and s. However, eq. (4) is remains a good approximation. There have been a few successful measurements of parity violation in the (n,y) reaction, where the total capture cross section was measured. In 139La an asymmetry A j of 9.5+/-0.3% has been measured[4]. Seestrom et al. [5] developed a neutron-capture detector, consisting of 24 Csl scintillators, for parity-violation studies at LANSCE. A measurement [6] of parity non-conservation in neutron capture on H1Cd and 113Cd observed large asymmetries, and a PNC mean-squared matrix element M=2.9+13_09 meV was obtained from the J=l levels in U4Cd. These sets of measurements showed that the (n,y)reaction could be used to obtain the same
205
level of information as the TRIPLE neutron transmission experiments, but on thinner targets. In the present paper we examine the advantages that can be gained by using high resolution gamma-detectors, allowing measurements of individual gamma-rays. Measurements of asymmetries for partial gamma-decays to individual final states would provide several measurements of A j from a given p-wave resonance. The requirement that the asymmetries be independent of the final state and all have the same value is a very strong test of the underlying theory. An observation of deviations from this prediction would imply that contributions from other neglected amplitudes may not be as small as previously assumed; there are some hints from the TRIPLE measurements that other reaction mechanisms may play a role. In the case of 232Th the measured asymmetries were all observed to have positive sign. The statistical nature of the compound nucleus, and the assumption that the neutron width amplitudes and PNC matrix elements are independent variables, makes theoretical interpretation of this common sign very difficult [7,8]. The assumptions made in deriving expression (4) for ALY are the same as those used in deriving expression (1) for ALn. The main approximation made is the assumption that the parity mixing all takes place when the neutron is captured into a p-wave resonance, i.e., that the />-wave resonance that is formed has a parity violating swave component. Within this approximation, once a neutron is captured into a parity mixed resonance the asymmetry arising from the o.k„ P-odd correlation no longer depends on the decay channel of the resonance. This assumes, for example, that there is no parity violation in the final state. While this is very physically reasonable, it cannot explain the sign correlation observed in 232Th. Here we are proposing a set of experiments measuring parity violation in partial neutron capture in both 232Th and in the better understood compound nuclei to resolve the issue. A measurement of the asymmetries in the (n,y) reaction for the same resonances studied by the TRIPLE collaboration would test the validity of the theory and shed light on the origin of the sign problem. An additional advantage of such measurements is that they could provide a more accurate determination of the parity mixing matrix element M, since several independent measurements of AL could be made for each resonance. 3
Experimental Considerations
We now turn to the question of the feasibility of these experiments. In the y-decay of the compound nucleus, the primary transitions of known multipolarity that can give information on PV are usually of high energy (E ~ 5-7 MeV), because they correspond to transitions from the capturing states to low-lying levels of known spin and parity. In contrast, the lower energy y-rays fall in the unresolved energy region of excitation, where no spectroscopic information is available on the individual energy levels.
206
Restricting ourselves to these higher-energy y-rays, the E\ transitions are on the average 7 times stronger than the Ml transitions [9], and several E\ transitions from a given p-wave resonance have been seen from resonances associated with enhanced parity-violating asymmetries. Partial radiation widths exhibit strong fluctuations as described by the Porter-Thomas distribution; thus, the relative intensity of specific £1 and Ml y-transitions of the similar energy can differ considerably from the average value of 7, and in some cases can be very large. In the case of many of the nuclei studied by the TRIPLE collaboration, e.g., 106 Pd, ,08Pd, 232Th and 238U, several E\ transitions from p-wave capture states to low-energy levels with opposite parity have been observed. The study of 139La would be more difficult, however, since all the low-energy states have the same parity as the p-resonances. Thus, only Ml transitions could be observed, making PNC measurements more difficult. Previously measured capture y-ray spectroscopy studies on /^-resonances give some indications on the feasibility of the class of experiments we are proposing. To explore the potential for parity violation studies using the (n,y) reaction, we consider the example of El and Ml y-rays from neutron capture on 107Ag, which have been studied at Geel [10]. Gamma-rays from several /"-resonances in the energy region of interest for PNC asymmetries [11] were studied. The emphasis in these experiments was on measurements of low-energy y-transitions, and thin samples were used to avoid absorption of low-energy y-rays from the sample itself. This meant that data in the high-energy region of the y-spectrum were available with good statistics for many s-wave resonances, but for only a few of the /?-wave resonances. Nevertheless, from these data we can estimate the ratio of partial radiation widths for a number of different pairs of £1 and Ml transitions. The absolute intensity of a transition can be obtained by dividing the measured number of observed counts by the sum of the intensities of all the transitions directly feeding the ground state and the isomeric states [12]. In Table I the absolute intensities of high-energy transitions from eight/>-wave resonances in the p +1C7Ag system are listed. Four of these /?-wave resonances, at 125.1, 259.9, 269.9 and 422.5 eV, exhibit PNC effects in transmission experiments, and all the observed ytransitions are of £1 character. We compare these with intensities of Ml transitions of the same energy from the close lying s-wave resonances with the same spin; our assumption being that parity mixing is dominated by mixing between neighboring opposite parity resonances. As can be seen from Table I, the observed ratio of transition intensities from/?-wave versus s-wave resonances can be as large as 100. In the cases where the Ml transitions from s'-resonances were not observed at all, despite the high statistics available for ^-resonances, the ryp/Tys ratio cannot be determined. Nonetheless, it is clear from Table I that the requirement for an enhanced PNC asymmetry ALY namely, that the condition ryp/Tys >1 is met. Then detailed knowledge of the partial gamma-widths is not necessary to extract a value of M2 from a set of measurements of Ai/1.
207 Table I Absolute intensities in photons per 100 neutron captures of high-energy transitions in 107Ag for neighboring p-wave and s-wave resonances with same spin. Statistical uncertainties only are indicated. The transitions from the p-wave (s-wave) resonances are E\ (Ml) in character. Several strong primary yrays from a given p-wave resonance are observed, suggesting that systematic studies of high precision PNC measurements may be possible.
EY(keV)
E 0 p-wave(eV)
i|P(o/0)
E 0 s-wave(eV) i*(°/0)
6450.4 64.2 0.36+0.05 51.6 6590.5 73.2 0.42±0.09 51.6 6690.4 0.37±0.06 6760.8 0.21±0.06 6803.5 0.35±0.09 6890.1 0.21±0.06 6760.8 107.6 0.68±0.15 51.6 6590.5 125.1 2.2±0.4 144.2 6803.5 183.5 0.37±0.06 202.6 6309.3 259.9 0.22+0.04 251.3 6504.0 0.23±0.04 7190.0 0.93+0.05 6803.5 269.9 0.44+0.08 251.3 6890.1 422.5 1.7+0.4 444.0 "Close to the single escape line from the 7269.4 keV transition.
0.0071±0.0012 Not observed 0.018±0.001 0.087±0.002a 0.16±0.002 Not observed 0.11±0.002a 0.022+0.005 0.003+0.001 Not observed 0.020±0.003 0.015±0.003 Not observed Not observed
We note that several other nuclei have been studied, and, in particular, similar results to the ones presented have been obtained for 232Th [13], proving that a measurement with a radioactive target is possible. As mentioned above, in the Geel measurements for />wave capture only the stronger primary transitions were observed. For systematic PNC studies, measurements optimizing the detection high-energy transitions are needed. In order to have sufficient statistics, it is important to have a high neutron flux at the measuring station, which needs to be at a sufficient distance to allow resolution of the resonances of interest. The main contribution to the degradation of the energy resolution in a spallation neutron source is the moderator. At the Lujan facility, for instance, to measure up to 500 eV requires a distance of 72 m. To optimize the count rate, the ideal measurements should (a) use a sample with higher mass than used in the Geel measurement, and (b) use germanium detectors with high efficiency. Additionally, a higher flux of neutrons in the energy range of interest that was available at Geel would be needed to maximize the number of observable y lines. In the cases of 232Th and 238U an additional difficulty arises from the fact that these nuclei involve radioactive targets. In this case it
208
would be preferable to operate with a small duty factor in order to reduce the background and increase the signal to noise ratio. 4
Summary
As noted by Flambaum and Sushkov, parity-odd correlations in radiative neutron capture can be very enhanced. Of the eight possible P-odd correlations that can occur in the (n,y) reaction we have concentrated here on the correlation a.k„. To a good approximation this leads to an asymmetry that is the same as the longitudinal asymmetry measured in neutron transmission experiments. A measurement of this correlation in total neutron capture cross sections have found asymmetries AL7 of the order of 10%. We emphasize that this asymmetry can be measured in the (n,y) reaction for several individual y-transitions from the same/?-wave resonance. Partial neutron capture measurements would allow high precision measurements of < VPNC>> and would provide an independent probe of important theoretical issues raised by the observations and analyses of the sign problem in 232Th. Finally, we note that systematic studies of parity violation in total (n,y) cross sections, would also provide a valuable test of the theory. References 1. J. D. Bowman et al, Phys. Rev. Lett. 65, 1192 (1990); X. Zhu et al., Phys. Rev. C 46, 768 (1992); C. M. Frankle et al Phys. Rev. Lett. 46, 1542 (1992); B. E. Crawford, et al, Phys. Rev. C 58, 1225 (1998). 2. V.V. Flambaum and O.P. Sushkov, Nucl. Phys. A435, 352 (1985). 3. V.E. Bunakov and V.P. Gudkov, Nucl. Phys. A401, 93 (1983). 4. T. Adachi et al, Nucl. Phys. A577 433c (1994). 5. S.J. Seestrom et al Nucl. Inst. Meth. A 433 (1999)603. 6. S.J. Seestrom et al, Phys. Rev. C 58 2977 (1998). 7. J.D. Bowman, G.T. Garvey, C.R. Gould, A.C. Hayes, and M.B. Johnson, Phys. Rev. Lett. 68,780,(1992). 8. A.C. Hayes and I.S. Towner, Phys. Lett. B 302, 157 (1993). 9. J. Kopecky and M. Uhl, in Measurement, Calculation and Evaluation of Photon Production Data, edited by C. Coceva, A. Mengoni an A. Ventura, Bologna, 1994. 10. L. Zanini, F. Becvar, F. Corvi, H. Postma, Phys. Rev. C 61, 054616 (2000). 11. L. Y. Lowie et al, Phys. Rev. C 59, 1119 (1999). 12. C. Coceva, Nuovo Cimento 107A, 85 (1994). 13. F. Gunsing, private communication, 2001.
TIME REVERSAL TESTS WITH EPITHERMAL NEUTRONS C.R. GOULD Physics Department, North Carolina State University, Raleigh, NC 27695-8202 and Triangle Universities Nuclear Laboratory, Durham, NC 27708-0308 We summarize the requirements for carrying out neutron transmission tests of time reversal violation and discuss the conditions under which the measurements will be competitive with bounds derived from electric dipole moments and other measurements. We review the issues related to extracting bounds from null measurements, and present the results of a first principles rederivation of the decomposition of the neutron-nucleus forward scattering amplitude for a target having arbitrary vector and tensor polarization moments.
1
Introduction
The factor of a million enhancement seen in parity violating observables for compound nuclear (CN) resonances and epithermal neutrons (see G.E. Mitchell, these proceedings) stimulated great interest in searching for time reversal violation (TRV). The enhancement derives from the close spacing and long lifetimes of CN resonances and is expected to be present at some level for all symmetry breaking observables in CN systems. For general background on the proposed experiments see Reft. [1,2]. To date no epithermal neutron test of TRV has been carried out for two reasons: difficulties in preparing a suitable target in which neutron precession does not average the effect to zero, and difficulties in controlling unwanted precessions associated with apparatus misalignments. Tests with higher energy neutrons have been carried out in holmium and the experimental precision can be high [3,4]. However, there are no CN enhancement mechanisms at work, and a 1/A suppression factor arises since only the last valence nucleon contributes to the TRV effect. Further improvement with MeV-energy experiments seems unlikely except in few nucleon systems where for example, a fivefold correlation measurement in pd scattering is being prepared for the COSY storage ring facility. Lacking more intense beams of MeV-neutrons, interest has therefore refocused on the practicality of epithermal TRV tests, and particularly at the SNS. In this paper we review the basics of TRV tests and consider two questions: 1.
Can neutron transmission tests of TRV compete with electric dipole moment (edm) measurements in setting bounds on TRV meson couplings (ft exchange for P-violating TRV, p exchange for P-conserving TRV). The answer is yes, although the experimental accuracies and beam intensities required are extremely challenging. Interestingly, there is one scenario for P-conserving TRV recently identified by Kurylov et al. [5], in which edm's set no bound at
209
210
2.
all. In this scenario (scenario B in their notation corresponding to parity symmetry not restored at the effective energy scale A of TRV) only direct measurements such as neutron transmission set reliable bounds. Assuming only null effects are seen, how many measurements are needed to set statistically significant bounds on TRV couplings. The answer is a minimum of two, but more practically, at least four to set a one sigma bound comparable to the experimental error in the measurements. This is more of a problem for Pviolating TRV tests than for P-conserving TRV tests due to the fewer number of resonances with known large parity violation. Of course a single non-zero result will be very exciting and therefore this argument should not be considered any reason for not proceeding ahead with experiments.
At the workshop a third possibility for studying TRV was discussed. These measurements would involve looking for a shift in the energy at which an (n, y) resonance capture asymmetry crossed zero. An early experiment of this type was carried out by Barabanov et al. [6]. It is not as yet clear that this proposed shift can be unambiguously disentangled from other effects. However, if the theoretical framework for analyzing the measurements can be confirmed, it does represent an additional intriguing avenue for pursuing TRV tests at the SNS. 2
Transmission Tests
Neutron transmission experiments search for symmetry violating terms in the neutron-nucleus forward scattering amplitude fe/{0). Heuristic considerations suggest that, for a polarized or aligned target, the dependence of/"e/(0) on the target spin / and the neutron spin s should be given by an expression of the form ([1] and references therein) fei(0) = fo+fMs-J
+ fpS-p +
fTs-{pxI),
where the complex coefficients f0, fu, fp and/?- are, in general, polynomials in I p . The symmetry terms of particular interest are: • •
$ •(! x p) - the P-odd, T-odd triple correlation (TC) with a polarized beam and a vector-polarized target with spin /. s '(Ixp)(I -p) - the P-even, T-odd fivefold correlation (FC), with a polarized beam and a spin aligned target (tensor polarization).
In general the scattering amplitude has real and imaginary parts, which respectively cause precession and attenuation, and there are three directions involved: p, I, and p x /. All terms can be acting simultaneously, including symmetry conserving terms like the strong interaction spin-spin term s •/. The resulting propagation of the beam
211 will be complicated. It is conveniently handled using the language of neutron optics, valid even for high-energy neutron beams as we are considering strictly forward scattering. The amplitude fej(0) and the refractive index n are 2 by 2 matrices in the spin space of the neutron, related by: n = 1 + (2n/l?)ffel(0). The time reversal tests TC and FC probe the symmetry of the S-matrix: ^LS.L'S' = 'is in the channel spin representation for example. The partial amplitudes fkKx making \xpfe$) ( s e e Appendix) are linear combinations of the anti-symmetric part of S when k + K + X is odd. Here k and K are the ranks of the polarization tensors describing the beam and target respectively, and X is the angular momentum transfer. For TC experiments, mixing of s- and p-wave amplitudes is required and the expectation is that the measurement will be carried out where parity violation is large. For FC experiments, mixing of s- and d-waves, or two separate p-waves is required. In contrast to parity violation where the underlying symmetry violating interaction is known, there is no accepted theory of time reversal violation. P-odd TRV can arise within the standard model but turns out to be second order weak, and therefore unobservably small. This is true for the neutron edm within the standard model also. In fact, a search for the TC term is analogous to a search for a neutron edm since a polarized target I represents an effective magnetic field B, and the resulting motional electric field E ~ (B x p) acts on the neutron to probe the presence of the same s • E interaction. Pion exchange is expected to dominate Podd TRV, and TC experiments are parameterized in terms of a T violating coupling constant gTK which from the neutron edm bound is expected to be < 2 x 10"11 [7]. P-even TRV does not occur in first order in the standard model, and its absence is a general feature of gauge theories with elementary quarks. The pion does not contribute to P-even TRV, and the longest-range contribution is due to p exchange or axial vector meson exchange [8]. Most of the experiments are parameterized by the ratio of the T-violating to T-conserving p couplings: gTp. The best bounds to date are from quite different experiments: • • •
FC studies of 6-MeV polarized neutron scattering from nuclear spin aligned 165 Ho at TUNL: gTp < 6 x 10-2 [4] The 199Hg edm bound: gTp < 1 x 10-2 [9] Charge symmetry breaking studies of 200-MeV n-p scattering at TRIUMF: gTp <0.7xl0" 2 [10]
Resonance tests have the potential to improve the bounds for both P-odd and Peven T-violation. Difficult experimental problems have to be overcome first however.
212 For the TC, the problems lie in a) accurately accounting for the effect of socalled sequential interaction terms, and b) identifying a sufficient number of cases in polarizable targets where parity violation is large. The effect of the TC term is equivalent to a "strong" rotation due to the spin-spin interaction term (pseudomagnetism) s • I followed by "weak" attenuation due to the helicity interaction term s -p. The action of these sequential terms is quadratic in the target thickness, in contrast to the linear dependence of the TC term, and they can also be detected by a polarizer-analyzer combination. Nevertheless, the expectation is that the search for a TC term will be carried out where parity violation is ten percent or more, and exceptional alignment of the apparatus will be needed to carry out a competitive measurement. For a recent discussion of the issues, see Ref. [11]. The observable of interest is the ratio of the TC cross section Om to the parity violating cross section doi at the resonance: RTP = Om/0101. Experiments are planned at KEK and in Dubna using frozen spin polarized lanthanum targets. An experiment competitive with the present edm limits will need to measure Rjp to order approaching 10 s [7]. For the FC, the problem lies in identifying suitable pairs of resonances with the same JF. The two possibilities are s-d or p-p mixing. Fortunately, there is a strong interaction signature - the deformation effect - that makes the s-d measurement especially attractive in 165Ho. The deformation effect arises from the term (/ • p)2 and is due to interference between the s- and d-wave amplitudes. In contrast to the (unmeasurably small) d-wave cross section, which scales with the d-wave width, the deformation effect cross section scales with the d-wave amplitude. It is an unambiguous signal of the presence of a d-wave component [12]. The s-wave resonance spacing is only 4 eV, and hundreds of resonances are known up to one keV or more. A nearby strong s-wave resonance with the same J1 is practically guaranteed by this high level density. As a result, multiple cases can be studied. An additional advantage of the FC experiment is that the sequential interaction problem is absent because the two terms that contribute are both P-odd: "weak" rotation due to the helicity term s • p followed by "weak" attenuation due to the deformation effect helicity interaction term (s • /)(/ • p). As a result, the measurement can be carried out without an analyzer. The observable of interest is a ratio of the FC signal to the deformation effect signal: Rm = FC/DE = 0122/0022- A two or more order of magnitude improvement may be possible at the SNS - see P. Huffman, these proceedings.
3
Statistics of Null Bounds
The complexity of compound-nucleus (CN) resonances necessitates a statistical description of on-resonance correlation cross sections. The first step towards a quantitative analysis of on-resonance FC (TC) measurements is the extraction of the local root-mean-square (rms) of CN matrix elements of a parity (P) conserving
213 (non-conserving) T-odd AW interaction F(T)(f/(PT)) or, in the event of null measurements, a bound on this root-mean-square. To this end, it is necessary to introduce a statistical reaction model which relates the unknown pure imaginary rms matrix element ( V ^ = ivT or V^P = i\PT) to the distribution of values of the onresonance correlation cross sections. In addition, one has to take into account the statistical errors in measurements. Since one is dealing with an intrinsically stochastic observable, the presence of errors can lead to non-trivial restrictions on the minimum number Mmin of on-resonance measurements required in effect to set a bound on vT and v^-. We have investigated [13] the statistics of null measurements of i?FD and RJP in a Bayesian framework. In the limit when mixing with only the nearest adjacent resonance is taken into account, simple closed-form expressions for the probability densities of /?FD and Rjp can be derived. Table I, summarizes the information for three standard choices of the probability or (corresponding to what for a gaussian random variable would be bounds at the 1, 2 and 3 <7 level, respectively). Because we work within a Bayesian framework, the bound v(a) on vT (vPT) has the immediate probabilistic interpretation that there is a probability a of vT(vPT) lying in the interval between 0 and v w .
v(«)
M
a = a2 a = a3 a = a° 7.12 2 >100. >5000. 80.4 3 1.53 10.3 4 0.786 3.77 17.4 5 0.521 2.12 7.68 0.387 1.44 4.52 6 7 0.308 1.08 3.09 0.255 0.860 2.31 8 0.712 1.83 9 0.218 1.50 10 0.190 0.607 0Ci == 0.683, oc2==0.955, oc3 = 0.997.
v(«)
a = ax
a = a2
« = «3
7.31 2.09 1.30 0.995 0.828 0.721 0.645 0.589 0.545
>100. 11.4 4.83 3.08 2.31 1.89 1.62 1.43 1.29
>5000. 81.3 19.8 9.56 6.10 4.48 3.57 2.99 2.59
Table I. Bounds v(a> on vT and VPT in units of the experimental error ac
One needs at least Mmin = 4 measurements to set a 68.3% probability bound on vT comparable in size to the experimental error, and at least Mmm = 5 to set a 68.3% bound on vpT. For M » 1, the asymptotic bounds on vT and VPT scale as (1/M) and (1/M)1/2 respectively. The unusual linear scaling with M for vT compared to vPT is similar to a result found by French et al. [14] in an analysis of TRV symmetry breaking in detailed balance studies. The more rapid decrease of bounds for vT compared to vPT doubly favors the FC experiments which are also more likely to
214 involve studying large numbers of cases. The prospects are therefore good for null FC measurements improving by two orders of magnitude or more upon the bound on ^ T ) extracted recently from neutron-proton charge symmetry breaking (CSB) experiments (currently, the tightest bound). It will be more difficult for a TC experiment to reduce the already stringent bound on 1^PT) from neutron and atomic edm measurements. Nevertheless, a single nonzero result would be a remarkable success and experimentalists should not be discouraged from proceeding in the execution of new experiments. In particular, a first result from a TC experiment on lanthanum would be most welcome. Acknowledgment This work was supported in part by US Department of Energy (Office of High Energy and Nuclear Physics) grant number DE-FG02-97ER41042. Appendix Despite the substantial reliance of analyses of transmission tests on the standard decomposition of fe/(0), there appears to be no complete formal derivation of the result in the literature. Instead, there are several presentations of a related decomposition of the total cross section (see [15] and the references therein). We recently undertook a first principles rederivation of the result [16] and we briefly summarize our findings in this appendix. Our full expression is: faak„,el(®)
=
/ JkKXskKl ' kKX
with partial scattering amplitudes LK
a
I
J
n
and polarization geometry correlation matrices sua
S
(-!)_
^ y - £ H ) '
V r i ^ V * )
q
••ik+K+X{{tK(I)®Cx(na)}k®Tk{s)}w. The cross-section coefficients are d
uJss-Sj±^^
L„SaLaSa
with angular momentum coefficients
215
J+ +
F$(L'S'IW
L
= (-l) * *^S'SL'L(L'OLO\AO)\ ' si
S' s I , * S s I S' J X k K L
[S
If the target nuclei have spin I = 1/2, our results imply that faaka,d(0) = a01 + aMPU2 • a + aPha -o + aT{ha
xPl/2)-a,
where, in terms of the partial scattering amplitudes fkxx, a
0 =/<)00 + / o i l w a '^1/2'
1 a
P
=_
7=/l01
3 H
a
M
=
. T=~fll2na'^\/2'
]7T^n2'
/ll0
1 a =
T ~r=f\\\
These results are consistent with the standard result once allowance has been made for the differences in notation. The results also agree with the total cross-section expression of Hnizdo [15] except we do not find the channel spin factor (-\f's that appears in his work. To our knowledge, explicit results of this form of the scattering amplitude have not previously been presented in the literature. References 1. "Tests of Time Reversal in Neutron Physics", ed. by Roberson, N.R., Gould, C.R. and Bowman, J.D. (World Scientific, Singapore, 1987) 2. "Time Reversal Invariance and Parity Violation in Neutron Reactions ", ed. by Gould, C.R., Bowman, J.D. and Popov, Yu. (World Scientific, Singapore, 1994) 3. Soderstrum, J.P. et al, Phys. Rev. C 38 (1988) 2424 4. Huffman, P.R et al, Phys. Rev. Rev C 55 (1997) 2684 5. Kurylov, A., McLaughlin G.C. and Ramsey-Musolf, M.J. Phys. Rev. D 63 (2001)076007 6. Barabanov, A.L. et al., Phys. Rev. Lett 70 (1993) 1216 7. Herczeg, P in "Symmetries and Fundamental Interactions in Nuclei", ed. by Haxton, W.C and Henley, E.M. (World Scientific, Singapore, 1995) p 89. 8. Beyer, M., Phys. Rev. C 48 (1993) 906 9. Haxton, W.C, Horing, A. and Ramsey-Musolf, M.J. Phys. Rev. D 50 (1994) 3422 10. Simonius, M. Phys. Rev. Lett. 78 (1997) 4161 11. Bunakov, V.E. and Novikov I.S., Phys. Lett B 429 (1998) 7 12. Huffman, P.R et al., J. Phys. G 24 (1998) 763
216 13. Davis, E.D. and Gould, C.R., Phys. Lett. B 447 (1999) 209. 14. French, J.B., Pandey, A. and Smith, J. in "Tests of Time Reversal in Neutron Physics", ed. by Roberson, N.R., Gould, C.R. and Bowman, J.D. (World Scientific, Singapore, 1987) p 80 15. Hnizdo, V., Phys. Rev. C 50 (1994) 2639. 16. Gould, C.R. and Davis, E.D. in "CP Violation and Related Processes", Lecture Notes in Physics, ed. by Beyer, M. (Springer-Verlag, to be published)
A N EXPERIMENT TO SEARCH FOR PARITY-CONSERVING TIME REVERSAL INVARIANCE USING EPITHERMAL N E U T R O N S FROM THE SPALLATION N E U T R O N SOURCE
P. R. HUFFMAN National Institute of Standards and Technology, Gaithersburg, MD 20899 We describe an experiment to search for parity-conserving, time reversal invariance in the neutron-nucleus cross section. This proposed technique will use the deformation effect in nuclear spin aligned 165 Ho to identify either p-wave resonances or small d-wave admixtures in predominantly s-wave resonances. If suitable resonances are identified, a test of parity-conserving, time-reversal violation in the neutron-nucleus cross section is possible, offering two or more orders of magnitude improvement in sensitivity over current limits.
1
Introduction
Neutron tests of time-reversal invariance have attracted considerable attention because of t h e large ( ~ 10 5 — 10 6 ) enhancements in sensitivity associated with compound nuclear resonances. Parity-violating asymmetries as large as 10 % have been measured in nuclei using b o t h neutron capture and transmission techniques 1 . Similar enhancements are expected for time-reversal asymmetries 2 , opening up the possibility for a sensitive test of time-reversal invariance. Mixing of t h e compound nuclear wave functions enhances t h e sensitivity to a time-reversal violating term. Tests of parity-conserving, time-reversal violation (PC TRV) have been proposed in t h e vicinity of two interfering resonances; enhancements exist if either two neighboring p-wave resonances interfere 3 ' 4 or a d-wave resonance interferes with a neighboring s-wave resonance 5 . We consider how to search for P C TRV at t h e Spallation Neutron Source (SNS) first using the deformation effect term in the the neutron-nucleus cross section as a signature of resonances suitable for a time-reversal test, t h e n performing t h e test on multiple resonance candidates. T h e particular test of time-reversal violation we are considering is t h e five-fold correlation (FC) which arises as a term in t h e neutron-nucleus cross section with t h e form frs • (I x k) I • k. Here s is t h e neutron spin, k is t h e neutron m o m e n t u m and I is t h e nuclear spin. T h e FC term can contribute if the neutron beam is polarized and the target is nuclear spin-aligned. In addition t o t h e F C term, t h e deformation effect (DE) t e r m - a symmetry conserving t e r m - is present when the target is spin-aligned and has t h e form
217
218 fo(I • k)2. Both the FC and DE amplitudes involve coupling the angular momentum of the neutron beam to the spin of the oriented target, with two units of angular momentum transferred. Neither the FC nor DE will be observable on a pure s-wave resonance. However, if the resonance is either a pwave or there is some d-wave admixture in an s-wave resonance, a deformation effect can be seen, and correspondingly there will be sensitivity to PC TRV mixing of these resonances with neighboring resonances. Thus locating either p-wave resonances or d-wave admixtures in multiple s-wave resonances opens up the possibility of a resonance time-reversal test. A more general version of this work has been published in Ref. 6. 2
Deformation Effect
General expressions for the polarized-beam, polarized-target cross section for isolated resonances are given in Refs. 5, 7, 8, 9. Following the notation of Ref. 9, the cross section can be written as a sum over the polarization ranks k of the beam and K of the target: o r = y^Jko{s)tKo{I)vkK
(1)
kK
with kK akK = 4 7 r A 2 ^ ? - I m V A C ^ A ( § i k ) V ] *J
A
(2J+l)£jj'
Jljl'j'
x (£A00|£'0) W(JjIK;
Ij'){
£ s j ) AkK) T$rtj. 'sj'\
(2)
Here, £fco(s) and txo(I) are the diagonalized statistical tensors describing the orientation of the beam and target respectively, T/,-,e- contains the Smatrix element describing the reaction, the angular momentum couplings are defined using the spin-orbit coupling scheme, j = £ + s, J = j + I, and d = (2a + l ) 1 / 2 . The geometrical factors are contained in the spin-correlation coefficients, CfcxA(sIk). The deformation effect arises from the term CT02 corresponding to an unpolarized beam (k = 0) and an aligned target (K = 2) ao2 = 47rA 2 ^!p 2 (cos0) V x <£200|£'0) W(JjI2;
( - l ) s - ^ " ( 2 J + l)ijj'
If) W(£j£'f; s2) Im {T^.tj
},
(3)
219 and the five-fold correlation arises from the term eri2(A = 2) corresponding to a polarized beam (k = 1) and an aligned target (K = 2) a 12 (A = 2) = 7 r A 2 ^ s i n 2 0 £ w
( - I ) - ' - * ' (2J +
l)£jf
£[£ +1) -£' (£' +1) - j(j + l) + f(f
+1)
>/*(* + 1) x <*200|*'0> W ( J j / 2 ; If) W(£j£'f; s2) Im { t T ^ } . Here 0 is defined as the angle between the directions of the diagonal tensors describing the target alignment and the beam direction. Note that the DE term has a P2(cos#) angular dependence and the FC has a sin 29 angular dependence. For convenience, we focus specifically on tests using 165 Ho. This is a nucleus which is monoisotopic and can be easily spin-aligned using cryogenic techniques. The deformation effect arises in holmium due to the non-spherical mass distribution of the nucleus. The spatial orientation of the nuclei in an aligned sample allows cross section measurements to probe different axes of the nuclei as the crystal alignment axis is rotated with respect to the beam direction. DE measurements were performed in holmium using transmission of MeV neutrons through a rotating, aligned 165 Ho target 10 . The 0° neutron yield is proportional to the total cross section and varies as P2(cos#) due to the DE (see Fig. 1). It is this oscillation that we plan to use to locate either p-wave resonances or d-wave admixtures in s-wave resonances. The deformation effect cross section at low energies on an isolated resonance is given by 6 (702(A) = TTA2 J g J P 2 (cos 9) {E _ E^J2
+ p 2 /4C(J*),
(5)
where gj = (2J + 1)/2(2J+1), and Ej and Tj are the energy and total width of the resonance of spin J. The coefficient C(JV) contains the neutron partial width amplitudes (<7^(^j)) along with their respective angular momentum coupling coefficients. Holmium has spin and parity of 7" = 7/2"" thus forcing p-wave resonances to have J71" = 2 + , 3 + , 4 + , 5 + and d-wave resonances to have J T = 1""... 6~. For an s-wave resonance with d-wave admixture, the deformation effect arises from terms involving £ = 0, £' = 2 and I = 2, I' = 0. (The £ = £' = 2 term will be negligible.) For a p-wave resonance, the deformation effect is due to the £ = £' = 1 term.
220 J1JVJ
3100
3050
it\r\r\
%%Au J
I
6950
I
I
I
I
I
I
I
7000
I
I
I
I
I
I
7050
Run Number Figure 1. The transmission yield for 9.4 MeV neutrons through 1 6 5 H o . Each run number corresponds to four minutes of d a t a taken in the angular sequence —180° -> +180° -» —180° in 22.5° steps. The oscillation arises from t h e deformation effect. Data taken from Ref. 10.
The size of the DE effect on an s-wave resonance with d-wave admixture or a pure p-wave resonance can be estimated using known level spacings, neutron widths, penetrabilities, and strength functions. Since weak resonances offer largest sensitivity to d-wave admixtures, estimates were made using a newly located 24.8 eV resonance in 165Ho(Ref. 11). These yielded ratios of lo'02/o'ool ~ 10~ 3 for both J w = 3~ or 4~. While it is unlikely that the resonance is p-wave, the deformation effect was estimated to be I002/C00I ~ 1, except for except J*' = 2 + , where |o"02/croo| = 0. Similar results were obtained for other low-energy weak resonances. The unpolarized cross section at 24.8 eV is 43 b. The small deformation effect associated with the d-wave admixture can be detected with a rotating aligned target and an unpolarized neutron beam. From the P2(cos#) angular dependence, the effect changes sign going from 0° to 90°, leading to a 0° —90° asymmetry. This asymmetry can be measured using either transmission or (n,7) capture techniques. The capture technique is more appropriate for studying weak resonances at the SNS because of the absence of a potential scattering background. This background has been shown
221 not to significantly contribute to parity violating asymmetries i 2 , and the same is expected for PC TRV measurements. Statistics limited measurements of both transmission and (n,7) radiative capture asymmetries with errors ~ 1 0 - 3 have been achieved at LANSCE 13,14 and KEK 12 . With the increased flux available at the SNS, weak d-wave admixtures, appear quite accessible to measurement at the SNS if systematic uncertainties can be controlled. Identification of p-wave resonances should be considerably easier. We note that the estimates of the DE are based solely on angular momentum barrier penetrability arguments, and do not assume any enhancement of d-wave amplitudes. It is well known that the tensor force mixes s-wave and d-wave amplitudes, and so there may well be enhancement of the d-wave components over what one might expect from simple penetrability considerations. This mixing has a dramatic effect on some low energy cross sections, enhancing for example, the 2 H(d, 7) 4 He capture cross section by orders of magnitude at energies of astrophysical interest 15 . We should also point out that large dwave admixtures in s-wave resonances have previously been observed in 2 0 7 Pb 16 . Resonances in the 100 keV - 300 keV energy range exhibited d-wave partial widths comparable to the s-wave widths (Tn(£ = 2) ~ Tn(£ = 0)). 3
Time-Reversal Violation for s-d Mixing
A 165 Ho resonance that exhibits a small deformation effect (
= -2^i9J
sin 20[{E^Ei)^{/A]
r2
M(J*),
(6)
where M{r)
= W{J\I2-
/ § ) gJnl (of) gJn2 (2§)
(7)
222
+2VQW(J\I2;ID
gJnl{0\)
gJn2(2l)
,
WT = \{1\HT\2)\ is the time-reversal violating matrix element coupling resonances 1 and 2, Y\ and 1^ are the total widths of the resonances, and the additional subscript on the neutron amplitudes gJni{(-j) indicates the ith resonance. The goal of the PC TRV measurement is to obtain a bound on Wrms, the root mean square value of the TRV matrix element WTComparison of the expressions for oyi andCT02shows them to be very similar. In particular, both scale linearly with the d-wave amplitude of resonance 2. It is therefore useful to consider the ratio of the cross sections, R, as the observable of interest. Substituting 6 — 45° into Eq. 7, we get
_ ^12(45°) -«7 02 (45°)
-2Wrmsw • z lrJnl(0) (E2~E1) Vr^2(0)
l
>
where w = WT/Wrms is a Gaussian random variable of unit variance and zero mean, and z is a second random variable containing both the neutron partial widths and angular momentum couplings 17 . The sensitivity of R to the PC TRV matrix element WrTns is determined by the probability distribution density of the product of the two random variables, w and z, and a constant C which is approximately 5 eV^ 1 for the case of the 24.8 eV resonance. The statistics of the random product w • z has been treated in detail in Ref. 17. Davis and Gould have shown that measurements on multiple resonances must be performed in order to extract meaningful bounds on Wrms. For holmium, measurements on at least four pair of resonances must be performed to extract a meaningful bound on Wrms. The strength of Wrms is conventionally expressed in terms of the ratio of the T-odd to T-even p-exchange coupling constants, denoted as gp. The most precise constraints on gp to date come from neutron-proton charge symmetry breaking (CSB) experiments, setting a 95.5 % confidence bound of gp = 6.7 x 10" 3 (Ref. 18). For this experiment, one must provide a bound of better than 50 meV for Wrms to be competitive with the CSB results 17 . Four null measurements of R = cri2/o"02 with an error of 0.1 would provide a 95.5 % bound of 40 meV and twelve measurements with the same accuracy would be about 4 meV 17 . Given the increased flux available at the SNS, one could expect to make a measurement of the asymmetry with a statistical accuracy of about 10~ 5 . This would imply that R could be measured to an accuracy of roughly 10~ 3 . With the minimum of four measurements, this would yield an improvement of two orders of magnitude on the limits of PC TRV. In the energy range accessible at the SNS, holmium has a few hundred s-wave resonances that
223
- 4ft detector
Figure 2. Schematic of a possible setup for the FC measurements.
could exhibit a deformation effect. In all likelihood, PC TRV measurements could be performed on many tens of resonances, yielding more than three orders of magnitude improvement over previous results. 4
Experimental Setup
A rough schematic of a possible experimental setup for performing a FC measurement is shown in Fig. 2. A complete discussion of false asymmetries can be found in Ref. 19. We focus specifically on the differences that would arise at the SNS. A polarized neutron beam (having passed through both a polarizer and spin flipper) enters the experimental region. The polarized neutrons pass through a beam monitor which is used to remove systematic uncertainties associated with beam current fluctuations. The neutrons then pass through the aligned 165 Ho target. The alignment axis of the target is in the plane perpendicular to the rotation axis. Neutrons which capture in the holmium are recorded by detection of the 7 emitted in the reaction using a ~ 4-7T segmented detector. The holmium target must be thin enough such that significant neutron depolarization does not occur. Measurements of neutron depolarization have been performed in 165 Ho and indicate that in our geometry, approximately 20 % of 1.7 eV neutrons depolarize when passing through a polarized cylindrical sample 2 cm in diameter 20 . Similar results were obtained for energies up to 37 eV. One must thus choose a thin sample (~ 1 mm) to minimize this effect. A double modulation technique of nipping the neutron spin while simul-
224
taneously rotating the alignment axis of the holmium target will be used to minimize systematic effects. Measurements taken at 6 = ±45°, ±135° will be used to extract the FC signal. Measurements taken at 6 = 0° and 180° can be used to search for false effects, since the FC term will be zero. 5
Summary
We have considered how to locate both p-wave resonances and d-wave admixtures in predominantly s-wave resonances in 165 Ho using the deformation effect. The rf-wave admixtures will be small and will most easily be seen in weak - as opposed to strong - s-wave resonances. If resonances can be found which have non-zero d-wave admixtures, then they are candidates for the fivefold correlation test of time-reversal. A measurement of the DE using (n,7) capture appears possible, which will open up the possibility of a sensitive test of time-reversal violation. Improvements of two or more orders of magnitude over the current best limits are possible. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
G. E. Mitchell et al., Phys. Rept. 354, 157 (2001). E. D. Davis and C. R. Gould, Phys. Lett. B 447, 209 (1999). V. E. Bunakov, Phys. Rev. Lett. 60, 2250 (1988). V. E. Bunakov et al, Phys. Rev. C42, 1718 (1990). C. R. Gould et al., Int. Jour. Mod. Phys. A5, 2181 (1990). P. R. Huffman, C. R. Gould, and D. G. Haase, J. Phys. G 24, 763 (1998). V. P. Alfimenkov et al., J. Nucl. Phys. 17, 149 (1973). A. L. Barabanov, Sov. J. Nucl. Phys. 45, 597 (1987). V. Hnizdo and C. R. Gould, Phys. Rev. C49, R612 (1994). P. R. Huffman et al., Phys. Rev. Lett. 76, 4681 (1996). P. R. Huffman et al., Phys. Rev. C54, 2051 (1996). H. M. Shimizu et al, Nucl. Phys. A552, 293 (1993). C. M. Frankle et al., Phys. Rev. C46, 778 (1992). E. I. Sharapov et al, in AIP Conf. Proc. 238 - Capture Gamma-Ray Spectroscopy, edited by R. W. Hoff (New York, 1990), p. 756. C. A. Barnes et al., Phys. Lett. B197, 315 (1987). D. J. Horen, J. A. Harvey, and N. W. Hill, Phys. Lett. 67B, 268 (1977). E. D. Davis and C. R. Gould, Phys. Lett. B 447, 209 (1999). M. Simonius, Phys. Rev. Lett. 78, 4161 (1997). P. R. Huffman et al, Phys. Rev. C55, 2684 (1997). V. P. Alfimenkov et al, Nucl. Instru. Methods A 352, 592 (1995).
NEUTRONIC CHARACTERISTICS OF THE SPALLATION NEUTRON SOURCE P. D. FERGUSON, E. B. IVERSON, AND F. X. GALLMEIER Spallation Neutron Source, Oak Ridge Nation Laboratory, Oak Ridge, TN, 37830-6474, USA E-mail: fergusonpd@ornl. gov A brief discussion of the characteristics, including the neutronic performance, of the SNS target system has been presented. Physical characteristics of the reflectors and moderators have been discussed and examples of the neutron energy spectra and time distributions have been presented. For detailed instrument design information, readers are directed to a website that will be kept up to date.
1
Introduction
The Spallation Neutron Source (SNS)[1], currently under construction at Oak Ridge National Laboratory, will be the most intense short-pulse spallation neutron source in the world when it is completed in 2006. Neutrons are created via the spallation process when the 1.4-mA, 1-GeV proton beam is incident on the liquid mercury target. After the neutron energy spectrum and time distribution are modified in the moderators and reflector, neutrons leave the target system along the 18 neutron flightpaths. An aerial view of the SNS site with renderings of the in-progress SNS structures is shown in Fig. 1. Typical short-pulse spallation sources involve four key components: 1) the ion source and front end, 2) the accelerator, 3) the proton storage ring/pulse compressor, and 4) the target system. One unique feature of the SNS project is the fact that each of these key components is being designed and constructed by one or more national laboratories according to the expertise available to the project. Instruments planned for the SNS include the study of materials and complex fluids, biological systems, and fundamental physics using cold neutrons among other things. However, several beamlines remain available should a compelling scientific program with identified resources be presented. The purpose of this paper is to document characteristics of the SNS that may be important to potential instrument design teams. In addition, references are presented to guide interested parties to detailed moderator leakage information critical to instrument performance calculations and optimization studies. In this paper, the SNS target system is described. The moderators and decoupling scheme are described in detail. Examples of the moderator energy spectra and time distributions are presented and a website is referenced where the complete set of calculated moderator performance files are stored.
225
226
Figure 1. Ariel view of the SNS construction site with an artist's rendering of in-progress and future structures.
2
SNS Target System
The SNS target system consists of a low-profile mercury target, a composite reflector, and a suite of four moderators. SNS is one of several high-power spallation sources in the design or construction phase with a target system based on a liquid mercury target [2,3]. Surrounding the mercury target is an inner reflector made of heavy water cooled beryllium with a nominal radius of 30 cm. Outside the beryllium is a stainless steel reflector (SS 304) with an outer radius of 100 cm. An elevation view of the target system is shown in Fig. 2. Above and below the target and inside the reflector are four moderators, three super-critical hydrogen and one light water. The moderator viewed surfaces for all moderators are 10 cm wide by 12 cm tall, although the moderating material extends beyond these dimensions. Above the target, the upstream moderator is a decoupled hydrogen moderator. The moderator is poisoned symmetrically at a poison depth of approximately 2.7 cm with an 0.8 mm thick gadolinium plate. The decoupling material on the moderator and along the flightpath in the inner reflector insert (radius of-50 cm) is cadmium. In the outer reflector insert (50 cm < r < 100 cm), the flightpaths of all moderators will be decoupled to reduce the long time tail from the neutron time distributions. Currently the decoupling material is cadmium, but discussion is underway to possibly change this decoupler to boron.
227
The top downstream moderator and the bottom downstream moderator are identical coupled hydrogen moderators. The moderators are 5.5 cm thick on average and are premoderated with a light water layer surrounding the hydrogen. As opposed to the two upstream moderators, the downstream moderators are viewed from only one surface. The bottom upstream moderator is an asymmetrically poisoned, decoupled light water moderator. As with the top upstream moderator, the decoupling material for the moderator and the inner reflector insert is cadmium. The poison material is also the same, gadolinium, but is slightly thicker (1 mm) to account for additional gadolinium depletion in the moderator over the estimated three-year life of the insert. For the water moderator, the gadolinium poison divides the moderator into a 2.5 cm thick moderator volume and a 1.5 cm thick moderator volume. The 1.5 cm thick volume is viewed from the downstream flightpath and the 2.5 cm thick volume from the upstream flightpath. The thinner moderating volume results in better energy resolution for neutron energies below the decoupling energy (-250 meV for cadmium) at the expense of neutron intensity. Estimates of the intensity loss for the 1.5 cm thick moderator volume are slightly less than a factor of two when compared to the 2.5 cm thick moderator volume.
Supercritical H 2 Moderator coupled
Supercritical Hj Moderator decoupled poisoned #
SS Reflector
Beryllium Reflector
Proton Beam
Ambient Moderator decoupled poisoned
Supercritical H 2 Moderator coupled
Figure 2. Isometric of the SNS target system.
3
Monte Carlo Modeling and Results
The neutronic performance of the SNS target system is predicted using the Monte Carlo program MCNPX [4]. One critical portion of the performance calculations is
228
the accurate geometric modeling of the target system. Figure 3 is a vertical slice through the MCNPX model of the target system, roughly corresponding to the position shown in Fig. 2. Because Fig. 3 is not a three-dimensional figure, the downstream hydrogen moderators do not appear (they are not located exactly in the center of the target system). The level of detail shown in Fig. 3 is believed to be accurate enough for the majority of instrument design and optimization studies. Additional detail is being added to the MCNPX target system model on a continual basis. A horizontal slice through the MCNPX model at the elevation of the bottom moderators is shown in Fig. 4, where the composite reflector structure and the physical details of the moderator, including the premoderator arrangement for the downstream moderators, are shown.
Figure 3. Elevation view of the MCNPX model of the SNS target system
229
decoupled H 2 0 moderator coupled LH2 moderator
SS reflector
reflector Figure 4. Plan view of the MCNPX model at the elevation of the lower moderator tier.
Performance calculations for the SNS target system typically imply neutron leakage calculations from each of the moderator viewed surfaces. From the MCNPX calculations, a well-converged calculation of the time-averaged neutron leakage energy spectrum is achievable in a reasonable time frame on an absolute basis. However, time distributions for discrete energy bin widths small enough for instrument design calculations are difficult to calculate on an absolute basis. The time-averaged calculations are generally used as the absolute scaling factor for the time distribution calculations, while the time distributions are calculated over a larger solid angle than instruments can physically view. The implicit assumption is that the time distribution does not vary greatly over the solid angles used for the calculations, an assumption that has been tested and is reasonable. As an example, the time-averaged energy spectra for the two viewed surfaces of the water moderator are shown in Fig. 5. As expected, the spectra are approximately the same for neutron energies above the cadmium decoupling energy. Below the cadmium decoupling energy, particularly around the thermalization peak, the spectra vary greatly with the thicker moderator, labeled High intensity H20, having a higher neutron flux. An example of the neutron time distributions is shown in Fig. 6, where the time distribution is shown for the two viewed surfaces of the SNS water moderator for 100 eV neutrons, and energy that is starting to become of interest for the ASAP collaboration. As expected from the time-averaged energy spectra, there is little difference in the intensity or time distribution from the different surfaces of the water moderator at energies above the decoupling energy.
230
ie+14 -
"1• A
High resolution H 2 0 High intensity H.,0
:
Neutron leakage (n/sr-pulse-eV)
1e+13 -
%
'ft 1e+12 -
:"
1e+11 -
\
1e+10 -
1e+9
-
0.001
0.01
0.1
1
10
100
Neutron energy (eV) Figure S. Time-averaged neutron leakage energy spectra from the two viewed surfaces of the SNS water moderator. 4e+10 •
• •
<> />
high resolution (1.5 cm poison) high intensity (2.5 cm poison)
3e+10
aj 6 £ a> £ to rag n o
100 eV neutrons
2e+10 *
il
•
4 4
1e+10
• • • * » , » « • 0.0
0.5
1.0
1.5
* 2.0
P 2.5
•»
' 3.0
Emission time (microseconds) Figure 6. Neutron leakage time distributions from the two viewed surfaces of the SNS water moderator for 100 eV neutrons.
231
One additional important assumption made for the SNS performance calculations is that protons arrive at the target at a delta function in time. For many calculations this is a good assumption. However, when the proton time distribution incident on the target becomes a significant fraction of the calculated neutron pulse width leaking from the moderator for the delta function proton pulse, then the proton time distribution incident on the target must be taken into account. The calculated proton time distribution extracted from the SNS proton storage ring is shown in Fig. 7. Accurate neutron time distributions for higher energy neutrons can be determined by folding the delta function calculations with the proton time distribution shown in Fig. 7. In the future, this detail will be account for in the distributed SNS performance calculations.
<
w c o
-400
-200
0
200
400
time (nsec) Figure 7. Calculated proton pulse time distribution extracted from the SNS proton storage ring.
Because of the large amount of data and the number of instrument design teams interested in the calculations, detailed performance calculations have been posted on the internet. The information can be found at the website: http://www.sns.anl.gov/components/moderators.shtml In addition to the calculations, there is also a document describing the calculations and an example showing one use of the information. If there is a need for information that is not available on the website, please contact the authors.
232
4
Summary
A brief discussion of the characteristics, including the neutronic performance, of the SNS target system has been presented. Physical characteristics of the reflectors and moderators have been discussed and examples of the neutron energy spectra and time distributions have been presented. For detailed instrument design information, readers are directed to a website that is kept up to date. 5
Acknowledgements
SNS is managed by UT-Battelle, LLC, under contract DE-AC05-00OR22725 for the U.S. Department of Energy. The authors would like to thank John Galambos, who provided the extracted proton pulse time distribution information. References 1. T. E. Mason, T. A. Gabriel, R. K. Crawford, K. W. Herwig, F. Klose, and J. F. Ankner, "The Spallation Neutron Source: A powerful tool for materials research," 20th International Linac Conference, August 21-25, 2000, Monterey, California, published on the LANL e-print server. 2. Y. Oyama, S. Ikeda, and JAERI-KEK Joint Project Team, "Status of Spallation Neutron Source Program in High Intensity Proton Accelerator Project," ICANS-XV 15th Meeting of the International Collaboration on Advanced Neutron Sources, Proceedings v. 1, Nov. 6-9, 2000, Tsukuba, Japan. 3. K. N. Clausen, "The European Spallation Source (ESS) Project," ICANS-XV 15th Meeting of the International Collaboration on Advanced Neutron Sources, Proceedings v. 1, Nov. 6-9, 2000, Tsukuba, Japan. 4. L. S. Waters, ed., "MCNPX Users Manual, Version 2.3.0," LA-UR-02-2607 (April 2002).
WORKSHOP SUMMARY: OPPORTUNITIES IN ASTROPHYSICS, SYMMETRIES, AND APPLIED PHYSICS AT SPALLATION NEUTRON SOURCES PAUL KOEHLER Oak Ridge National Laboratory CHRISTOPHER GOULD North Carolina Stale University and TUNL ROBERT HAIGHT Los Alamos National Laboratory TIMOTHY VALENTINE Oak Ridge National Laboratory The presentations and discussions of the workshop are briefly summarized. As a result of the workshop, an Instrument Development Team is being developed and a Letter of Intent (Lol) has been submitted to the nascent Spallation Neutron Source at Oak Ridge National Laboratory. Because the Lol summarizes much of the science discussed at the workshop, it follows this brief summary in lieu of a typical, longer summary.
Several of the many programs of research in astrophysics, symmetries, and applied physics at existing neutron facilities were described and discussed at this workshop. These programs have a long and distinguished history and continue to be important and productive today. The need for many new experiments and the exciting opportunities they present were discussed. A common thread in many of these new opportunities is the need for higher neutron flux in the epithermal energy range. To this end, the new Spallation Neutron Source (SNS) being built at the Oak Ridge National Laboratory will have the highest flux of pulsed epithermal neutrons in the world when it comes on line in 2006. As a result of the workshop, an Astrophysics, Symmetries, and Applied Physics Instrument Development Team (ASAP IDT) has been assembled and a Letter of Intent (Lol) has been submitted to the SNS. The ASAP-Lol currently is under review by the SNS. Because this Lol summarizes much of the science discussed at the workshop, it follows this brief summary in lieu of a typical, longer summary. If the ASAP-Lol is approved by the SNS, then a full proposal to the SNS and funding agencies will be developed and submitted. Including ASAP and its state-of-the-art apparatus at the SNS will enable many new exciting and important experiments in the fields of nuclear astrophysics, the study of the violation of fundamental symmetries, and applied nuclear physics.
233
LETTER OF INTENT TO THE SPALLATION NEUTRON SOURCE Instrument Team title and acronym: Astrophysics, Symmetries, and Applied Physics (ASAP). Spokesperson:
Paul Koehler, Senior R&D Staff Physics Division Oak Ridge National Laboratory MS-6354, Bldg. 6010 Oak Ridge, TN 37831 Tel: 865-574-6133, Fax: 865-576-8746 E-mail: koehlerpe(S;ornl.gov
Team members: C. Alexander1, J. Andrzejewski2, C. Baktash1, D. Bardayan1, J. Batchelder , J. Blackmon', C. Brune4, M. Busso5, K. Carter3, A. Champagne6, V. Cianciolo1, D. Clayton7, A. Couture8, Y. Danon9, F. Difilippo1, E. Esch10, P. Ferguson1, W. Furman11, K. Furutaka12, R. Gallino13, Y. Gledenov", E. Gonzalez-Romero14, S. Goriely15, C. Gould16, G. Greene10, U. Greife17, S. Grimes4, V. Gudkov18, G. Gueorguiev19, R. Haight10, J. Harvey1, H. Harada12, A. Hayes10, M. Herman20, N. Hertel21, P. Huffman22, M. Igashira23, C. Jordan7, F. Kappeler24, G. Kim25, R. Kozub26, K.-L. Kratz27, A. Laptev28, L. Leal1, J. Lynn10, T. Massey4, Y. Masuda29, A. Mengoni30, B. Meyer7, G. Mitchell16, P. Mueller1, Y. Nagai31, J. Neal1, P. Oblozinsky32, T. Osaki23, S. Penttila10, F. Rahnema21, S. Raman1, W. Rapp33, T. Rauscher34, J.-P. Renier1, B. Rundberg10, R. Sayer1, H. Schatz35, O. Shcherbakov28, M. Smith1, K. Toth1, J. Ullmann10, T. Valentine1, S. Wender10, M. Wiescher8, J. Wilhelmy10, R. Winters36, G. Young1, V. Yuan10. 'Oak Ridge National Laboratory, 2University of Lodz, Poland, 3Oak Ridge Associated Universities, Ohio University, 5Osservatorio Astronomico di Torino, 6University of North Carolina, 7Clemson University, 'University of Notre Dame, 'Rensselaer Polytechnic Institute, ,0Los Alamos National Laboratory, "Joint Institute for Neutron Research, Dubna, Russia, 12Japan Nuclear Cycle Development Institute, Tokai-mura, Japan, 13University of Torino, Italy, 14CIEMAT, Madrid, Spain, lsUniversite Libre de Bruxelles, Belgium, l6North Carolina State University, ''Colorado School of Mines, "University of South Carolina, l9 University of Florida, ^International Atomic Energy Agency, Vienna, Austria, Georgia Institute of Technology, 22National Institute of Standards and Technology, Tokyo Institute of Technology, Japan, Forschungszentrum Karlsruhe, Germany, 25Kyungpook National University, South Korea, 26Tennessee Technological University, "University of Mainz, Germany, 28Petersburg Nuclear Physics Institute, Russia, 29High Energy Accelerator Research Organization (KEK), Japan, 30ENEA, Bologna, Italy, ''Research Center for Nuclear Physics, Osaka University, Japan, 32Brookhaven National Laboratory, "University of Karlsruhe, Germany, 34University of Basel, Switzerland, ^Michigan State University, 36Denison University 4
234
235
Scope of Work 1
Introduction
We propose to use the intense flux of epithermal (=0.5 eV to 300 keV) neutrons at the Spallation Neutron Source (SNS) to perform world-class experiments in nuclear astrophysics, symmetries, and applied nuclear physics. Because the SNS will be by far the most intense pulsed source of epithermal neutrons in the world, new vistas of unexplored areas will be opened in each of these fields. For example, it will be possible to use much smaller samples and thereby make measurements on rare and radioactive isotopes. In addition, the higher counting rates at the SNS will be a boon to basic physics experiments by allowing improved precision, better exploration of possible systematic uncertainties, and new experimental approaches. Research in nuclear astrophysics will concentrate on experiments to determine the astrophysical reaction rates for radioactive samples needed to interpret recent astronomical observations and to improve models of nucleosynthesis occurring in stars and supernovae, of galactic chemical evolution, and of the formation and age of our solar system. Research in symmetries will be aimed at exploiting, at epithermal energies, the unique properties of the interactions between neutrons and nuclei to study properties of many-body quantum systems such as the transition from purely statistical to coherent states, to study the violation of fundamental symmetries such as parity and time reversal invariance, and to better determine basic properties of the neutron such as its electric polarizability. In applied physics, the emphasis will be on measuring neutron cross sections for radioactive isotopes needed to address important issues in nuclear reactor physics, the transmutation of nuclear waste, criticality safety, stockpile stewardship, material analysis, and the improvement of nuclear models. The importance of the basic physics component of our proposed research program has been emphasized in reports generated during the recent long-range planning exercise of the Nuclear Science Advisory Committee (NSAC). For example, the report from the town meeting on Nuclear Structure and Astrophysics concludes that there are "exciting opportunities in nuclear astrophysics at the future high flux Oak Ridge Spallation Neutron Source" and that a "program should be implemented using the intense neutron beams from the SNS for establishing a longrange future facility for nuclear astrophysics using neutrons." In the report from the similar town meeting on Astrophysics, Neutrinos, and Symmetries, improvements in symmetry experiments made possible by the intense flux at the SNS were noted. The importance of this research to applied nuclear physics has been underscored by several national and international organizations such as the International Atomic Energy Agency in their request lists detailing high priority nuclear data needs. Another example of the importance of the applied nuclear physics program is the recent Department of Energy Defense Programs announcement of a Stewardship
236
Science Academic Alliances Program that specifically called for proposals addressing "low energy cross sections of stable and unstable nuclei...for neutroninduced reactions for radiochemistry diagnosis". Such measurements require the use of very small radioactive targets, so the high flux available at the SNS will be extremely valuable in this endeavor. 2
Proposed Research Program
On March 11-13, 2002 we held a workshop at the Joint Institute for Heavy Ion Research in Oak Ridge on Astrophysics, Symmetries, and Applied Physics at Spallation Neutron Sources. Approximately 70 scientists from almost 30 institutions around the world were in attendance. The aim of this workshop was to review current and near-future research in these fields, and to begin to assemble the team of researchers and the scientific case for future research at the SNS. Almost all the attendees of the workshop as well as several scientists who could not attend have joined this letter of intent and plan to be part of the future Instrument Development Team (IDT). The scientific case for the SNS in these fields is summarized below. 2.1
Astrophysics
Nuclear reactions involving neutrons play vital roles in many astrophysical scenarios. For example, virtually all elements heavier than iron were made in environments inside stars and supernovae or more exotic environments where neutron interactions dominate the nucleosynthesis. Neutrons dominate because charged particle rates are highly suppressed by large Coulomb barriers and because astrophysical conditions favor the release of large fluxes of neutrons. In addition, many of the most interesting new astronomical results, from the latest generation of observatories to measurements of isotopic anomalies in meteorites, are providing views of the results of this nucleosynthesis with unprecedented detail and precision. Furthermore, more realistic models of astrophysical environments made possible by recent advances in numerics and computing are providing new insights into the inner workings of these objects, as well as contributing to related topics such as galactic chemical evolution and the formation and age of our solar system. However, further progress in these areas is hampered by the lack of accurate rates for nuclear reactions governing nucleosynthesis. Many of these astrophysical reaction rates can be determined by measuring neutron-induced cross sections in the energy range between approximately 1 eV and 300 keV. One example of a key reaction rate that cannot be measured at any other facility is Ni(«,/?) Co. Observations with the Compton Gamma-Ray Observatory have detected radioactive 57Co from the recent supernova SN1987A. Such observations are potentially very valuable diagnostics of supernova explosions if the nucleosynthesis of this isotope can be understood. In a recent study of the
237
nucleosynthesis of Co, it was determined that only a single rate, Ni(w,p) Co, was crucial for improving the precision of the predicted 57Co abundance. Due to the short half life of 57Ni, the measurement would be very challenging. But, scaling from previous experiments using the calculated 5 Ni(n,p)57Co cross section indicates that a measurement should be feasible at the SNS but not at any other facility. In most cases, progress in nuclear astrophysics requires more global improvements in the nuclear data. In particular, reaction rate measurements are urgently needed for radioactive isotopes, for stable isotopes of very low natural abundance, for low neutron energies (below 5 keV), and for isotopes with small cross sections. Due to flux limitations at current facilities, very few of these types of measurements have been made. An examination of the observed elemental abundances in the solar system, together with rudimentary nuclear physics considerations, reveals that there were three main processes responsible for the origin of the elements heavier than iron. Almost all these elements are thought to have been synthesized inside stars, supernovae, or other more exotic environments through sequences of neutron capture reactions and P decays during the so-called slow neutron capture (or "s") and rapid neutron capture (or "r") processes. The s and r processes are each responsible for roughly half of the observed heavy element abundances. The remaining neutron-deficient isotopes that cannot be reached via neutron capture pathways are thought to have been formed in massive stars or during supernova explosions through the photodissociation, (or "p") process. In the s process, the neutron density is fairly low (~108 cm") and hence the reaction path follows the valley of 3 stability, and the resultant abundance of a given isotope is, in general, inversely proportional to its («,y) reaction rate. Because the reaction path follows the valley of P stability, many of the required («,y) reaction rates have been measured, and the mean properties of the s process environment are thought to be fairly well understood. However, for about 20 radionuclides along the s-process path, the neutron-capture and P-decay time scales are roughly equal. The competition between neutron capture and P decay occurring at these isotopes causes branchings in the reaction path that can be exploited to directly constrain dynamical parameters of ^-process models. Elucidating these dynamics should lead to important insights about the inner workings of stars. At present, many dynamical variables are essentially free parameters in the astrophysical models, because there are almost no data on the (n,y) reaction rates for radioactive branching points. To learn about the dynamics, accurate (~5%) rates for the relevant neutron capture reactions are needed so that the relative strengths of the neutron-capture and p-decay paths in the branchings can be calculated. A key dynamical ingredient of sophisticated s process models is the treatment of complicated mixing occurring between various layers inside red giant stars. Similar mixing processes are important in various stages of stellar evolution and in almost all nucleosynthesis environments. Because the nuclear physics information needed for s' process models is the most
238
accessible to experiments, studying the s process offers perhaps the best hope for obtaining a better understanding of mixing in astrophysical environments. Because radioactive samples are challenging to make and work with, a technique requiring minimum sample size is highly favored. The flux at the SNS is expected to be over 10,000 times larger than at the ORELA facility where most previous («,y) measurements have been made and approximately a factor of 40 higher than at LANSCE. Therefore, it should be possible at the SNS to measure many important samples not accessible at other facilities. Similarly, the large flux at the SNS should make possible for the first time (n,y) measurements on many radioisotopes of interest to the r and p processes. Overall, measurements should be possible on about 50 radioactive isotopes at the SNS. We have made a prioritized list of the required measurements by examining the most likely reaction pathways in the most recent and successful astrophysical models. Neutron capture measurements at low energies (En=0.5 eV to 5 keV) and on very rare stable isotopes, and («,oc) measurements are additional areas of nuclear astrophysics where the large flux available at the SNS should be of great value. For example, very few («,y) measurements have been made on stable isotopes having very small natural abundances because the required samples are too expensive for current techniques, and very few («,cc) measurements have been made because the cross sections are very small. These measurements will be particularly important for obtaining a better understanding of the p process. At present, the p process is so poorly understood that the astrophysical site or sites where it occurs is unknown, although the late stages of nuclear burning in massive stars or supernova explosions appear to be the leading candidates. 2.2
Symmetries
Research in this area will concentrate on measurements of symmetry violations and on physical properties of the neutron itself. The substantial boost in flux at the SNS compared to current facilities will make possible the next generation of experiments in these areas. Enhancement effects in low energy neutron resonances can increase parity violating (PV) observables by as much as a factor of 106. As a result, the TRIPLE collaboration has measured =80 statistically significant PV asymmetries in compound nuclear resonances with their apparatus at LANSCE. Due to the complicated structure of compound nuclear resonances in heavy nuclei, PV effects are interpreted using statistical approaches. A basic tenet of this approach is that there should be, on average, equal numbers of negative and positive PV asymmetries. One puzzling result of the TRIPLE data is that all ten of the PV asymmetries measured for 232Th below En=250 eV have the same sign. At present, this result is ascribed to some poorly understood nuclear structure effect. The TRIPLE experiments concentrated on the regions of nuclear masses where the
239
effects were expected to be strongest and on isotopes where fairly large samples could be obtained. At the SNS it would be possible to extend these measurements to higher energies, to other mass regions, and to smaller samples. In addition, by using the same detector needed for nuclear astrophysics, it would be possible to use even smaller samples (and hence a wider variety). These extensions are crucial for obtaining a better understanding of properties of many-body quantum systems, in particular the transitions from pure statistical to coherent states. Neutron induced reactions are unique for these studies because symmetry properties of nuclei scale as the neutron energy, which is negligible compared to the energy of the formed compound nucleus. Therefore, by performing experiments at neutron energies up to 100 keV, it is possible to make substantial changes in the symmetry properties of the system without significantly perturbing other system properties. Because individual resonances will not be resolved in these "continuum-averaged" experiments, enhancements of PV observables are expected to be smaller than in previous studies of individual resonances. Therefore, these experiments would be very difficult if not impossible at current facilities, but the larger flux at the SNS will make them possible. Via these experiments it should be possible to study the statistical properties of quantum systems and their transitions from pure statistical (chaotic) to coherent (self-organized) states. In particular, neutron induced fission experiments are expected to be very rich laboratories for these types of studies as well as for the study of many unsolved problems in models of nuclear fission. All these experiments would allow better tests of the assumptions of PV models and a better understanding of possible nuclear structure effects. Also, both of these achievements may have an important impact on possible searches for time-reversal invariance violation (TRIV) experiments using compound nuclear resonances. TRIV is another example of a symmetry violation that could be enhanced in heavy nuclei. Although the CP-violation observed experimentally in the decay of K° mesons implies TRIV, it has yet to be observed directly without additional theoretical assumptions. The observation of a direct TRIV effect would have major implications in a number of areas. By exploiting the possible enhancements due to neutron resonances, it may be possible to observe TRIV either in parity-violating (Podd, as is the case for K° decay) or parity-conserving (P-even) interactions. Such experiments require very high statistical precision as well as a careful study of systematic effects through ancillary experiments. The high flux available at the SNS makes it the best facility in the world for these experiments. There are at least two properties of the neutron that require epithermal neutrons for their measurement. Although both of these properties - the electric polarizability and the charge radius of the neutron - reportedly have been measured, there is still significant controversy concerning the results. It should be possible to determine both of these quantities via high-precision measurements of the neutron total cross section for lead in the few eV to few keV region. The high flux, high peak intensity,
240
and low gamma background available at the SNS make significant improvements with respect to previous experiments possible, and should allow the definitive measurements of these quantities to be made. A definitive measurement of the polarizability of the neutron has been of high interest because the result could have a significant impact on quark models of the nucleon. 2.3
Applied Physics
By enabling cross section measurements on the widest range of radioactive samples, the SNS offers substantial advantages over current facilities for experiments of importance to nuclear power, the transmutation of nuclear waste, criticality safety, stockpile stewardship, material analysis, and the improvement of nuclear models. There has been much recent interest in, and a significant increase in funding for, studies into extending the lives of the current generation of nuclear reactors and in concepts for a new generation of reactors. Several of the proposed reactors have designs that differ significantly from conventional nuclear power plants. As a result, new fission and («,y) cross section measurements are needed on a number of actinides and fission products to assess the feasibility of these non-standard reactor concepts. Advanced designs include fuel loadings with higher initial enrichments of uranium, mixed-oxide fuels, tighter lattice pitch, and longer refueling intervals. These differences alter the neutron spectrum in the core and result in significant changes in the neutron reaction rates in the core as compared to conventional nuclear power plants. The longer refueling intervals are of particular interest because these result in the production of more fission products due to the higher burn-up of the reactor fuel. The impact of the fission products on the decay heat produced in the core, the reactivity feedback coefficients, and the margin of excess reactivity required for startup must be examined. Fission products that are not too important in conventional reactor designs may become more important as refueling intervals become longer. Additionally, the longer refueling intervals will increase the production of the actinides that will also impact these issues. Studies assessing the feasibility of transmuting nuclear waste require similar measurements. For example, («,y) cross sections are needed for several fission products, and («,y) as well as fission measurements are needed for actinides. Obtaining samples for and overcoming backgrounds from radioactive samples will be a challenge in many of the needed measurements. The large flux at the SNS will make possible a large reduction in the sample sizes needed and hence the widest range of these measurements. Specific isotopes that are too difficult for current facilities but should be measurable at the SNS include 79Se, 135Cs, 238Pu, 241Am, 232 U,and242'243'244Cm. Measurements of («, y) cross sections for radioactive samples are also of particular interest for nuclear criticality safety. The Nuclear Regulatory Commission (NRC) currently permits the nuclear industry to take credit for depletion of uranium
241 and actinides that impact the criticality of spent nuclear fuel in storage and/or transport. In the future, the nuclear industry and the NRC both envision taking allowance for the production of fission products that significantly reduce the reactivity of spent nuclear fuel because of the large capture cross sections of the fission products (burn-up credit). The implementation of full burn-up credit will be dependent upon the adequacy and availability of cross section data for the fission products. Without adequate data, the industry cannot take full benefit for the reduction in reactivity caused by the presence of the fission products. The interpretation of radiochemical data from past nuclear weapons tests requires accurate cross section data for reactions on radioactive nuclei which are intermediate products in the explosion. Over 60 reactions on radionuclides have been identified as being important, and many of these are (n,y) reactions at energies below 300 keV. It is planned that the DANCE detector being built at LANSCE will be used for some of these measurements; however, many will remain outside the capability of that instrument, but several can be studied with the higher neutron flux available at the SNS. Neutron capture resonance analysis is another possible application of the proposed instrument. With the high flux at the SNS it should be possible to quickly determine the presence of many different elements in a given sample with extremely high selectivity. Finally, there will always be neutron capture, fission, and other cross sections of importance to basic and applied science that are beyond the reach of even the best available techniques. For these cases, we must rely on nuclear models. By making new measurements off the valley of p stability, the SNS will lead to improvements in these models so that they are more reliable when extrapolated to immeasurable cases. Experimental Approach We propose to instrument a beam line viewing the water moderator at the SNS, because this moderator is expected to have the best resolution at epithermal energies. A crucial aspect of many of the proposed experiments will be the ability to resolve resonance structure over the widest range of energies. A standard figure of merit that has been used for many years to assess the ability to resolve resonances is the flux divided by the square of the time resolution, FOM = (|>/(At/t)2. At 50 eV (the highest energy for which we could find calculations), the intrinsic figure of merit for the water moderator is calculated to be at least 5 times better than, for example, the coupled hydrogen moderator. Also, several of the anticipated experiments will be particularly sensitive to backgrounds from y rays; hence, enough shielding will be needed around the detectors that it does not appear possible to use of one of the "dual-channel" beam lines (e.g. beam line 8). In addition, several of the experiments would benefit from the better resolution available on a relatively long (-100 m)
242
flight path. Therefore, the most suitable location for our apparatus appears to be beam line 7 or 9. There should be no negative impact on the neutron scattering capabilities of the SNS. The heart of the instrument for most of the anticipated experiments will be a 4n y-ray detector, most likely made from the scintillator BaF2, at a flight path distance of about 30 m. No neutron guide will be needed, but excellent collimation will be required to allow the highest flux and minimum beam halo at the sample position in the center of the detector. The detector will be highly segmented so that event rates in the individual elements are manageable and to aid in background reduction. Charged particle detectors for fission, (n,p), and (w,cc) measurements will be placed in the beam at a well shielded location in front of the y-ray detector. These components will be based on the latest detector (Parallel Plate Avalanche Counters, Compensated Ion Chambers) technology. The data acquisition system will take advantage of latest advances in digitizing electronics. It is anticipated that it will be possible to run fission, (n,p), or («,oc) measurements simultaneously with neutron capture experiments with the BaF2 detector. These digital electronics can distinguish between types of events (in this case gammas or scattered neutrons) at the expected high rates through the use of flash adcs and peak shaping software. The collaboration will build on the acquisition systems already in place by the nuclear structure group at HRIBF (ORNL), and the nTOF collaboration at CERN. This suite of detectors will enable the astrophysics and applied physics experiments and will be used for many of the symmetries experiments. The addition of a neutron polarizer, very likely at a position just outside the bulk shield, and a spin flipper will be required for most of the symmetries experiments. Also, some symmetries experiments will require an aligned or polarized sample, a different neutron polarizer, and other apparatus. We plan to locate this apparatus behind the BaF2 detector to maximize beam usage. The symmetries program could be expanded, and some basic nuclear physics experiments would benefit from a longer flight path on the order of 100 m. Although we do not expect to implement this longer flight path as a part of the initial instrument, we request that this option not be precluded. ASAP Collaboration The multi-institutional ASAP collaboration is being formed to bring together the expertise and resources needed to take advantage of the unique opportunities at the SNS for world-class research in nuclear astrophysics, symmetries, and applied nuclear physics. Team members from US and foreign universities and national laboratories have extensive experience in these areas of research at the ORELA and LANSCE white neutron sources as well as other facilities. For example, members from Karlsruhe and LANL have designed, operated, and are in the process of
243
building new, large 4% y-ray detectors. Team members from ORNL are experienced in applying the latest digital data acquisition systems to complex experiments and members from LANL are developing similar apparatus. Team members at ORNL, LANL, and Colorado School of Mines have many years of experience and access to the specialized equipment needed for producing radioisotopes and fabricating radioactive targets. Team members from Dubna, Karlsruhe, Lodz, and ORNL are experts in designing, constructing and using detectors for charged particles at white neutron sources. It is expected that students from collaborating universities will make substantial contributions to this effort. Many of the anticipated experiments, which we estimate will require one to two weeks of data taking, should make excellent dissertation topics. The neutron polarizer and other major pieces of the apparatus used in the TRIPLE experiments at LANSCE are still world class, and could be moved to the SNS for PV and TRIV experiments. Several of our team members from North Carolina State University, LANL, and KEK designed and are experienced in operating this apparatus as well as instrumentation for polarized and aligned samples. A unique polarized lanthanum target system is also potentially available from KEK. Finally, our team members include theorists who have many years of experience in developing and testing astrophysical and nuclear models using the results from experiments such as those we plan to undertake at the SNS. The collaboration will be led by a spokesperson and we expect to form three teams each containing specialists in the three principal areas of research being proposed. Construction and operation of the instrument will be the responsibility of Physics Division at ORNL, which has extensive experience operating and supervising the research at user facilities. The user activities will be coordinated and managed by Oak Ridge Associated Universities, which has a long history of facilitating university research at national laboratories. We expect outside users to contribute substantially to the design, construction, and usage of the facility and these users will include a large number from both domestic and foreign universities. Mode of Access -IDT We are forming an IDT because the proposed research does not involve neutron scattering and because construction and operating funds will be derived from sources outside SNS. Based on recent experience of members of our team in constructing a beam line at LANSCE as well as the involvement of some of us in the design of a beam line for an IDT for fundamental physics experiments at the SNS, we estimate that the construction of the collimation and shielding will cost $1M. Based on our experience constructing similar detectors at Karlsruhe and LANSCE, the design and construction of the 4n y-ray detector needed for many of the experiments is expected
244
to cost approximately $2M. We estimate that moving the TRIPLE apparatus from LANSCE, and installing it at the SNS will require ~$200k. The ionization chambers and their associated electronics should require about $150k. Operation of this instrumentation will require about 2-3 FTE of effort. It is anticipated that several program offices at the Department of Energy (Nuclear Physics, Defense Programs, Nuclear Energy) will have programs using the proposed instrument and thus may participate in funding its construction and operation. There appears to be a high probability of obtaining the needed funding. Conclusion Our proposed ASAP instrument will use the intense flux of epithermal (=0.5 eV to 300 keV) neutrons at the Spallation Neutron Source (SNS) to perform world-class experiments in nuclear astrophysics, symmetries, and applied nuclear physics. New vistas of unexplored areas will be opened in each of these fields by coupling the most intense pulsed source of epithermal neutrons in the world to state-of-the-art instrumentation. With ASAP it will be possible to use much smaller samples and thereby make cross section measurements on rare and radioactive isotopes - an area largely unexplored at present. In addition, the higher counting rates will be a boon to basic physics experiments by allowing improved precision and better explorations of possible systematic uncertainties, and by enabling new experimental approaches.
AUTHOR INDEX Barabanov, A. 184 Blackmon, J.C. 146 Bowman, J.D. 58, 155 Chanberlin, E.P. 16 Derrien, H. 115 Eschm, E.-I. 58 Ferguson, P.D. 225 Fowler, M.M. 16 Fujii,T. 131 Furman, W. 184 Furutaka, K. 131 Gallmeier, F.X. 225 Gonzalez, E. 83 Gould, C.R. 209,233 Guber,K.H. 97,115 Gudkov,V. 194 Haight,R.C. 16,233 Harada,H. 131 Hayes, A.C. 202 HeiLM. 16 Herman, M. 107 Huffman, P.R. 217 Hunt, L. 16 Igashira, M. 52 Iverson, E.B. 225 Jordan IV, G.C. 42 Kaeppeler, F. 1, 16 Katoh,T. 131 Kobayashi, K. 131 Koehler, P.E. 32,233
Laptev, A.B. 123 LeaLL.C. 97,115 Lynn, J.E. 65 Makii,H. 52 Masuda,Y. 175 Matthews, J. 58 Mengoni, A. 25 Meyer, B.S. 42 Miah,M.M.H. 131 Mishima, K. 52 Mitchell, G.E. 155 Morgan, G. 58 Nagai,Y. 52 Nakamura, S. 131 O'Donnell, J.M. 58 Oblozinsky, P. 73 Ohsaki,T. 52 Penttila, S.I. 155, 164 Popov, A. 184 Reifarth, R. 16 Rundberg, R.S. 16 Seabury, E. 16 Segawa, M. 52 Sharapov, E.I. 155 Shcherbakov, O.A. 123,131 Shima, T. 52 Spencer, R.R. 115 Strottman, D.D. 16 Tomyo, A. 52 Ullmann, J.L. 16
245
246
Valentine, T.E. 97,233 Vorobyev, A.S. 123 Wender, S.A. 58 Wilhelmy, J.B. 16
Wright, R.Q. 115 Yamana, H. 131 Yuan, V.W. 138 Zanini, L. 202
The spallation neutron source (SNS) being built at the Oak Ridge National Laboratory (ORNL)will be by far the highest flux pulsed source of epithermal neutrons in the world when it comes on line in 2006. Although the main thrust of the science program at the SNS will be materials science, the facility could provide outstanding opportunities for research in nuclear astrophysics, fundamental symmetries, and applied nuclear physics. To review the current status of these fields and to begin to assemble the scientific case and the community of researchers for future experiments at the S N S, a workshop on "Astrophysics, Symmetries, and Applied Physics" was held in March 2002 at the ORN L. Over 60 scientists, representing 11 US and 4 foreign universities as well as many national laboratories around the world, participated in the workshop. The proceedings describe the current state of research in those fields and the future opportunities at the SNS.