197Au(n,γ) cross section in the resonance region

PHYSICAL REVIEW C 81, 044616 (2010) 197Au(n,γ) cross section in the resonance region C. Massimi,1C. Domingo-Pardo,2,*G. Vannini,1L. Audouin,3C. Guerrero,4U. Abbondanno,5G. Aerts,6H. ´ Alvarez,7 F. ´ Alvarez-Velarde,4S. Andriamonje,6J. Andrzejewski,8P. Assimakopoulos,9,†G. Badurek,10 P. Baumann,11 F. Beˇ cv´ aˇ r,12 F. Belloni,5E. Berthoumieux,6F. Calvi ˜ no,13 M. Calviani,14 D. Cano-Ott,4R. Capote,15,16 C. Carrapic¸o,6,17 P. Cennini,14 V. Chepel,18 E. Chiaveri,14 N. Colonna,19 G. Cortes,20 A. Couture,21 J. Cox,21 M. Dahlfors,14 S. David,3I. Dillmann,22 W. Dridi,6I. Duran,7C. Eleftheriadis,23 L. Ferrant,3,†A. Ferrari,14 R. Ferreira-Marques,18 K. Fujii,5W. Furman,24 S. Galanopoulos,25 I. F. Gonc¸alves,17 E. Gonz´ alez-Romero,4F. Gramegna,26 F. Gunsing,6B. Haas,27 R. Haight,28 M. Heil,22 A. Herrera-Martinez,14 M. Igashira,29 E. Jericha,10 F. K¨ appeler,22 Y. Kadi,14 D. Karadimos,9D. Karamanis,9M. Kerveno,11 P. Koehler,30 E. Kossionides,31 M. Krtiˇ cka,12 C. Lampoudis,32 C. Lederer,33 H. Leeb,10 A. Lindote,18 I. Lopes,18 M. Lozano,16 S. Lukic,11 J. Marganiec,8S. Marrone,19 T. Mart´ ınez,4P. Mastinu,26 E. Mendoza,4A. Mengoni,14,15 P. M. Milazzo,5 C. Moreau,5M. Mosconi,22 F. Neves,18 H. Oberhummer,10 S. O’Brien,21 J. Pancin,6C. Papadopoulos,25 C. Paradela,7 A. Pavlik,33 P. Pavlopoulos,34 G. Perdikakis,25 L. Perrot,6M. T. Pigni,10 R. Plag,22 A. Plompen,32 A. Plukis,6A. Poch,20 J. Praena,26 C. Pretel,20 J. Quesada,16 T. Rauscher,35 R. Reifarth,28 M. Rosetti,36 C. Rubbia,14 G. Rudolf,11 P. Rullhusen,32 L. Sarchiapone,14 R. Sarmento,17 I. Savvidis,23 C. Stephan,3G. Tagliente,19 J. L. Tain,37 L. Tassan-Got,3L. Tavora,17 R. Terlizzi,19 P. Vaz,17 A. Ventura,36 D. Villamarin,4V. Vlachoudis,14 R. Vlastou,25 F. Voss,22 S. Walter,22 M. Wiescher,21 and K. Wisshak22 (n TOF Collaboration) 1Dipartimento di Fisica, Universit` a di Bologna, and Sezione INFN di Bologna, Italy 2GSI Helmholtzzentrum f¨ ur Schwerionenforschung GmbH, Darmstadt, Germany 3Centre National de la Recherche Scientifique/IN2P3-IPN, Orsay, France 4Centro de Investigaciones Energeticas Medioambientales y Tecnologicas, Madrid, Spain 5Istituto Nazionale di Fisica Nucleare, Trieste, Italy 6CEA/Saclay-IRFU, Gif-sur-Yvette, France 7Universidade de Santiago de Compostela, Spain 8University of Lodz, Lodz, Poland 9University of Ioannina, Greece 10Atominstitut der ¨ Osterreichischen Universit¨ aten, Technische Universit¨ at Wien, Austria 11Centre National de la Recherche Scientifique/IN2P3-IReS, Strasbourg, France 12Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic 13Universidad Politecnica de Madrid, Spain 14CERN, Geneva, Switzerland 15International Atomic Energy Agency (IAEA), Nuclear Data Section, Vienna, Austria 16Universidad de Sevilla, Spain 17Instituto Tecnol´ ogico e Nuclear-ITN, Lisbon, Portugal 18LIP-Coimbra & Departamento de Fisica da Universidade de Coimbra, Portugal 19Istituto Nazionale di Fisica Nucleare, Bari, Italy 20Universitat Politecnica de Catalunya, Barcelona, Spain 21University of Notre Dame, Notre Dame, Indiana, USA 22Karlsruhe Institute of Technology, Institut f¨ ur Kernphysik, Germany 23Aristotle University of Thessaloniki, Greece 24Joint Institute for Nuclear Research, Frank Laboratory of Neutron Physics, Dubna, Russia 25National Technical University of Athens, Greece 26Istituto Nazionale di Fisica Nucleare, Laboratori Nazionali di Legnaro, Italy 27Centre National de la Recherche Scientifique/IN2P3-CENBG, Bordeaux, France 28Los Alamos National Laboratory, New Mexico, USA 29Tokyo Institute of Technology, Tokyo, Japan 30Oak Ridge National Laboratory, Physics Division, Oak Ridge, Tennessee, USA 31NCSR, Athens, Greece 32EC-JRC-IRMM, Geel, Belgium 33Faculty of Physics, University of Vienna, Austria 34Pˆ ole Universitaire L´ eonard de Vinci, Paris La D´ efense, France 35Department of Physics-University of Basel, Switzerland 36ENEA, Bologna, Italy 37Instituto de F´ ısica Corpuscular, CSIC-Universidad de Valencia, Spain (Received 29 January 2010; published 27 April 2010) 0556-2813/2010/81(4)/044616(22) 044616-1 ©2010 The American Physical Society C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) The (n,γ) cross section of 197Au has been measured at n TOF in the resolved resonance region, up to 5 keV, with the aim of improving the accuracy in an energy range where it is not yet considered standard. The measurements were performed with two different experimental setup and detection techniques, the total energy method based on C6D6detectors, and the total absorption calorimetry based on a 4πBaF2array. By comparing the data collected with the two techniques, two accurate sets of neutron-capture yields have been obtained, which could be the basis for a new evaluation leading to an extended cross-section standard. Overall good agreement is found between the n TOF results and evaluated cross sections, with some significant exceptions for small resonances. A few resonances not included in the existing databases have also been observed. DOI: 10.1103/PhysRevC.81.044616 PACS number(s): 25.40.Ny, 25.40.Lw, 29.30.Hs, 27.80.+w I. INTRODUCTION The main objectives of the experimental activity of the neutron time-of-flight facility, n TOF, at CERN, are accurate measurements of neutron cross sections related to nuclear astrophysics [1,2] and the collection of nuclear data related to emerging nuclear technologies for energy production and nuclear-waste transmutation [3–5]. Most neutron cross sections are measured relative to crosssection standards [6] for normalization to absolute values. So far, the 197Au(n,γ) reaction at thermal energy and between 0.2 and 2.5 MeV is the only capture standard and most neutroncapture cross-section measurements refer to one or both energy regions. An alternative to the use of cross-section standards is the saturated resonance technique [7] using a low-energy saturated resonance, like, for example, the 4.9-eV resonance in a197Au(n,γ) reaction. Owing to its high capture cross-section value, this resonance is saturated for a sample thickness greater than 30 µm. Because of the convenient neutron-induced radioactivity, chemical and isotopic purity, and large thermal neutron capture and resonance capture integral, the Au capture cross section is of great importance, for example, for flux measurements in nuclear reactors, in accelerator mass spectrometry, and in neutron-activation analysis. The 197Au(n,γ) cross section is not very accurately known in the resolved resonance region (RRR). The few previous measurements were carried out with liquid scintillation detectors containing H or F and did not cover the full RRR up to ≈5 keV. Resonance parameters up to 1 keV were determined in Refs. [8–10] by combining the results of different types of neutron cross-section measurements (i.e., transmission through thick and thin samples, capture, self-indication, and elastic scattering) and using the so-called area analysis [11]. From the resonance shape analysis of a transmission measurement Alves et al. [12] determined resonance parameters from 1 to 2.5 keV. In the energy region 2.5–5 keV, capture data from Macklin et al. [13] were combined with differential elastic scattering data from Hoffman et al. [14]. The evaluated cross sections in the neutron-reaction libraries ENDF/B-VI [15] and ENDF/B-VII [16] (the latter based on the compilation of Ref. [17]) show small discrepancies. In addition, a few resonances reported by Desjardins et al. [8] and Julien et al. [10] are only partly included. *Corresponding author: [email protected] †Deceased. This situation motivated a new measurement of the capture cross section of 197Au at the n TOF facility with the aim of establishing the Au capture standard also in the energy range below 200 keV. To reduce systematic uncertainties as far as possible, the measurement was carried out with different gold samples and by using two independent detection techniques based on a total absorption calorimeter (TAC) and a pair of C6D6detectors (Sec. II). This article presents the results of a resonance shape analysis with the R-matrix code SAMMY [18] for the resolved resonances in the energy region between 1 eV and 5 keV. The analysis procedure for the TAC and C6D6data is illustrated in Secs. III and IV, respectively. The comparison of the two data sets with each other and with evaluated cross section data is given in Sec. V. The unresolved resonance region between 5 keV and 1 MeV is being analyzed in parallel and will be presented separately [19]. II. MEASUREMENTS A. The n TOF facility During Phase I of the n TOF facility (2001–2004) the neutron beam was produced by spallation induced by a 20 GeV/c proton beam, with up to 7 ×1012 particles per pulse, impinging on a 80 ×80 ×60 cm3lead target with a repetition rate of 0.4 Hz. These characteristic features of n TOF allow one: (i) to cover the neutron-energy interval from 1 eV to 250 MeV in a single run, (ii) to achieve an extremely high instantaneous neutron flux, and (iii) to prevent pulse overlap even for subthermal neutrons. A 5.8-cm-thick water layer surrounding the lead target serves as coolant and as a moderator of the initially fast neutron spectrum, providing a wide neutron energy spectrum with a nearly 1/Enisolethargic flux dependence in the neutron energy region from 1 eV to 1 MeV. An evacuated beam line leads to the experimental area at a distance of 185 m from the lead target. The neutron beam line is extended for an additional 12 m beyond the experimental area to minimize the background from backscattered neutrons. A full description of the characteristics and performance of the facility can be found in Refs. [20–22]. The neutron beam is shaped by two collimators at 135 and 175 m from the spallation target. For capture measurements, the second collimator is used with an inner diameter of 1.8 cm, 044616-2 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) resulting in a nearly symmetric Gaussian-shaped beam profile at the sample position with a standard deviation of about 0.77 cm at low neutron energies [23]. The neutron energy is determined via time of flight (TOF), using the γflash from the impact of the proton pulse on the spallation target as the time reference. The relationship for converting TOF into neutron energy was accurately verified in the energy range from 1 eV to ∼1 MeV by means of specific capture resonances in 32S, 193Ir, and 238U, which are accepted energy standards [24]. For each detector signal, the corresponding TOF is determined on an event-by-event basis with an accuracy of about 2 ns. The data acquisition system (DAQ) [25] with 54 channels consists of high-frequency flash analog-to-digital converters (FADCs) [26]. Each channel has an 8-Mbyte memory buffer and is operated at a rate of 500 Msamples/s. In combination with the low duty cycle, the DAQ allows one to record the full sequence of signals in each detector in a TOF interval from relativistic neutron energies down to approximately 1 eV. This operation mode corresponds to a zero dead-time data acquisition that is important for avoiding large dead-time corrections at low neutron energies, where the (n,γ)cross section of Au is rather large. After zero suppression, the data are reduced and stored in the CERN central data recording system. Specially designed pulse-shape-analysis routines are used in the data-reduction stage to extract amplitude, integrated slow and fast component, and TOF from the digitized detector signals. This information, together with the corresponding detector number and the number of protons in the respective pulse, are then used for further data analysis. For more details, see Ref. [25]. B. Neutron-capture detectors Neutron-capture events are characterized by γ-ray cascades leading from the excited state to the ground state of the compound nucleus formed in the reaction. In the n TOF measurements, a total-energy-detection system with two C6D6 liquid scintillation detectors, as well as a total γ-ray-absorption calorimeter (TAC), have been used for measurements of capture cross sections. These two techniques are briefly described in the following. A first set of measurements was carried out using two C6D6detectors, which have been specially designed [27] with the aim of reducing the γ-ray background induced by neutrons scattered in the sample and captured in or near the detectors. As illustrated in Refs. [28,29], this background has been recognized as a relevant source of error in previous measurements. Recorded events in the C6D6detectors need to be treated by the pulse height weighting technique [30]to achieve the proper energy-dependence of the γ-ray efficiency, as described in more detail in Sec. IV. The n TOF TAC [31–35]isa4πdetector with nearly 100% detection efficiency for capture γ-ray cascades and an energy resolution of 15% at 662 keV and 6% at 6.1 MeV. It consists of 40 BaF2crystals contained in 10B-loaded carbon-fiber capsules forming a spherical shell 15 cm in thickness and with an inner diameter of 20 cm. Neutrons scattered from the sample in the center of the TAC are moderated and partly absorbed in a TABLE I. Gold samples for the two capture measurements. TAC C6D6 Diameter (cm) 1.0 2.205 Mass (g) 0.1854 1.871 Thickness (cm) 1.22 ×10−22.5×10−2 Areal density (at/b) 7.3×10−41.498 ×10−3 5-cm-thick spherical shell made of C12H20O4(6Li)2surround- ing the sample. The TAC is ideal for capture measurements of low mass samples, as well as of radioactive and fissile isotopes, owing to its very high total efficiency and because it allows one to select capture reactions via the total energy of the γ-ray cascade and to reject events attributable to other processes, in particular in-beam γrays from neutron captures in the water moderator of the spallation target. A certain drawback of the device is the relatively high neutron sensitivity, mostly attributable to the capture of scattered neutrons in the Ba isotopes of the scintillator (Sec. III B2). To some extent this problem has been reduced by means of the absorber shell around the sample and the 10B-loaded carbon-fiber capsules. Contrary to the TAC, the C6D6setup is optimized for cases where the total cross section is strongly dominated by the elastic channel. These detectors are, in fact, characterized by a very low neutron sensitivity of about 10−4, two orders of magnitude smaller than that of the TAC, thus providing reliable results even for very small γ/n ratios. The setup for the capture measurements is complemented by the silicon flux monitor (SiMon) [36], which consists of a thin 6Li deposit on a thin Mylar foil surrounded by a set of four silicon detectors outside the neutron beam for recording the tritons and αparticles from the 6Li(n,α)3H reaction. C. Samples Gold samples, which differed in size and thickness, were used in the measurements to control sample-related systematic effects. The characteristics of the samples are listed in Table I. In addition to gold, samples of natC and natPb of the same diameter as the Au samples have been used to evaluate the background owing to sample scattered neutrons and in-beam γrays. III. ANALYSIS OF THE TAC DATA The energy calibration of each individual BaF2crystal was obtained by means of standard γ-ray sources, that is, 137Cs (662 keV), 88Y (898 and 1836 keV), and Pu/C (6131 keV from 16O). The energy resolution of each BaF2module and of the entire array were obtained from these measurements as well. A. From measured count rate to capture yield The processed information from the 40 BaF2crystals is combined off line in a so-called calorimetric routine with the aim of identifying capture events. Although the time resolution of each crystal is less than 2 ns owing to the very fast decay time of the BaF2scintillators, the overall time resolution of the TAC 044616-3 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) is larger (i.e., 26 ns) owing to the uncertainty in the calibration and synchronization of the different FADCs. Therefore, the condition that defines an event in the TAC is the recording of signals in two or more crystals within a coincidence window of 26 ns. A 150-keV threshold is used for the individual signals to reject electronic noise and to minimize pileup effects (see Sec. III B1). For each processed event, the calorimetric routine returns the total energy deposited in the TAC (ETAC ), the incoming neutron energy (En), and Mγ, the number of BaF2 crystals in which a γray is detected above threshold. The segmentation of the TAC is enough to ensure a close correlation between the multiplicity of the detected event and the number of γrays emitted in the capture cascade. The probability that a capture reaction occurs in the sample is the capture yield, that is, that fraction of neutron beam that undergoes a capture reaction in the sample. Experimentally, it is obtained from the ratio of the total counts detected by the TAC, CAu(En), and the incoming neutron fluence (En) integrated over the beam profile, Yexp(En)=CAu(En)−Cempty(En) ·f·(En),(1) where Cempty(En) are the counts measured without the sample and represent the sample-independent background (other sources of background are discussed later), is the TAC efficiency for detecting a capture event, and fis the fraction of the neutron beam intercepting the sample. The correction factors and fare independent of neutron energy in the range considered here. The efficiency depends on the conditions of the analysis, that is, on the multiplicity window and the energy cuts chosen for the TAC response. The fact that the neutron beam profile varies very slightly with neutron energy was also properly taken into account. Because the absolute normalization in our analysis is obtained via the 4.9-eV saturated resonance, it is not necessary to know the absolute value of the flux, but only the relative energy dependence of the neutron flux up to few keV, which has been measured with a 235U parallel plate fission ionization chamber of PTB Braunschweig, Germany [37]. At neutron energies below 1 keV, we used the flux from the SiMon, which was normalized to the former data in the overlapping energy region. After the normalization factor Nwas determined via the 4.9 eV resonance, the experimental capture yield is Yexp(En)=N·CAu(En)−Cempty(En) (En).(2) Figure 1shows the total energy deposited in the TAC for the samples used in the measurement. The peak at 6.5 MeV corresponding to the excitation energy of the compound nucleus formed after a neutron capture on 197Au is clearly visible. Moreover, background components are also present. A delicate part of the data analysis consists of the choice of the optimal thresholds for the deposited energy ETAC to maximize the capture-to-background ratio. The selection criteria in the present analysis are illustrated in Figs. 1and 2. The adopted conditions are 3.5<E TAC < 7.5 MeV on the total deposited energy and Mγ⩾2. As shown 10-3 10-2 10-1 1 10 2 4 6 8 10 12 14 ETAC (MeV) counts/proton bunch H(n,γ) Au(n,γ) Ba(n,γ) B(n,αγ)n + Au n + empty sample n + C FIG. 1. (Color online) The spectra of the energy deposited in the TAC measured with the Au and C samples compared to the case without sample (empty) in the neutron energy range 1 < En<5000 eV. The adopted thresholds for the deposited energy are indicated by dashed vertical lines. later, the choice for ETAC minimizes the neutron sensitivity, because it allows us to reject the 2.2-MeV γrays produced by hydrogen capture in the absorber around the sample completely and neutron captures by the Ba isotopes in the crystals partly (in particular from the odd nuclei 135,137Ba, which are characterized by capture energies above 7 MeV). As a further advantage, pileup of two consecutive capture cascades, which mimics events with large total energy deposition, is reduced (Sec. III B1). Although the overall efficiency decreased to 60% by these conditions, the resulting signal-to-background ratio is drastically improved, as shown in Fig. 2. B. Corrections and background evaluation The capture yield measured with the TAC must be corrected for systematic effects, before performing a resonance analysis 10 -4 10 -3 10 -2 10 -1 1 1 10 10210 3104105106 En (eV) Capture yield No conditions 3.5<ETAC<7.5 MeV FIG. 2. (Color online) Experimental capture yields for the 197Au(n,γ) reaction extracted with and without the selection criteria for the deposited energy ETAC . 044616-4 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) with an R-matrix code. The corrections are related to the dead time of the detector setup and to its neutron sensitivity. 1. Pileup and dead time The large counting rate associated with the very high instantaneous neutron flux at n TOF results in two different effects that have to be considered in the analysis of the TAC data, that is, pileup of consecutive signals in each BaF2crystal and the occurrence of two consecutive capture events within the 26-ns coincidence window used in the calorimetric routine. In analogy to standard electronics, we will refer to this second effect as detector dead time. In principle, the n TOF DAQ should not be affected by pileup because the FADCs allow one to identify and to reconstruct two consecutive signals (see for example Ref. [38]), contrary to standard electronics and acquisition systems. However, a correction is still required if two signals are too close in time to be correctly identified by the reconstruction routine [31], especially if a small signal occurs on the tail of a preceding larger signal. In this way, a fraction of low-energy γrays may be lost, thus distorting primarily the multiplicity and, to a lesser extent, the total energy deposited in the TAC. The corresponding correction is based on the exact signal shape of each crystal and on the ability of the reconstruction routine to identify pileup events. One possibility is to rely on detailed simulations of the detector response as shown in Ref. [31]. Another approach, which has been adopted in the present study, takes advantage of the fact that the nTOF proton beam is delivered in two different modes, a dedicated mode with an intensity of 7 ×1012 protons/pulse and a parasitic mode with approximately half the intensity. The comparison between data collected in both modes reveals that pileup problems are affecting only the low-energy part of the energy deposited in each crystal. Therefore, the pile-up effect is reduced to a negligible level by an energy threshold of 150 keV for the individual crystals and by the multiplicity condition (Mγ⩾2). The second, more important effect of the high counting rate at n TOF is related to the occurrence of two capture events within the coincidence window used in the calorimetric routine to sum up all γrays belonging to a capture event. For the largest observed counting rates (in the range 1–3 counts/µs), the probability of detecting two capture events in the coincidence window cannot be neglected. As shown in Fig. 3, an increase of the time window from 26 to 52 ns results in a reduction of the capture yield (obtained with the present analysis conditions on ETAC and Mγ). Owing to the constraints on total deposited energy, the combination of two (or more) capture reactions leads to the loss of one or both events, depending on whether the resulting ETAC falls within the adopted pulse height window of 3.5<E TAC <7.5 MeV. The effect is analogous to the loss of counts due to the dead-time in standard data processing and acquisition systems. Therefore, the counting rate for this detector dead time has been corrected in first order by means of the standard noncumulative (nonparalyzable) model (see for example, Ref. [39]) with a dead time corresponding to the width of the coincidence window. 0 0.02 0.04 0.06 0.08 0.1 0.12 3.5 4 4.5 5 5.5 6 6.5 En (eV) Yield not normalized 26 ns time window 52 ns time window FIG. 3. (Color online) Capture yields extracted with coincidence time windows of 26 and 52 ns to illustrate the detector dead time at the example of the first Au resonance at 4.9 eV. In the present analysis a time window of 26 ns was adopted. The ETAC spectra in parasitic and dedicated mode for the TOF region around the resonance at 4.9 eV are compared in Fig. 4. Pileup at the higher count rate in dedicated mode is clearly visible by the enhancement beyond 9 MeV. The standard assumption that out of two or more coincident events only one is detected represents a first-order correction. For the calorimetric method, however, other cases should be considered as well. With the conditions selected for ETAC and Mγ, both events can be lost if the sum falls outside the limits of the total deposited energy. However, some events, which were lost for dead time, would have been lost anyway, because they did not match the analysis conditions initially. The true number of lost events was estimated by means of the Monte Carlo (MC) method. In the simulation, two events were randomly chosen from the measured ETAC spectrum and added together. The resulting spectrum is shown in Fig. 5together 10-6 10-5 10-4 10-3 10-2 02468101214 ETAC (MeV) counts/total counts Parasitic mode Dedicated mode FIG. 4. (Color online) The ETAC spectra of 197Au for parasitic and dedicated operation mode for capture events around the 4.9-eV resonance. 044616-5 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) 1 10 10 2 10 3 10 4 10 5 0 2 4 6 8 10 12 14 16 ETAC (MeV) counts Exp. data MC simulation Pile up FIG. 5. (Color online) Experimental distribution of the energy deposited in the TAC for neutron capture events around the 4.9-eV resonance in Au and a Monte Carlo simulation of the pileup effect for capture events taken from the same distribution. with the experimental ETAC distribution. The second-order correction was determined by simulations of pileup events. In particular, we have calculated the probability that pileup events fulfill the analysis conditions for two randomly chosen events from the experimental distribution. To avoid the background from sample-scattered neutrons in the ETAC spectrum, the simulation was performed in the energy range between 4.8 and 5.1 eV. It was found that the resulting probability for pileup corresponds on average to a 20% increase of the dead-time correction. Accordingly, the counting rate was corrected by the following expression: C r=Cr 1+Crτ(F2−1) 1−Crτ,(3) where C rand Crare the corrected and recorded count rates, respectively, τis the dead time (i.e., 26 ns), and F2 is the second-order correction (1.2 on average). The main uncertainty of the dead-time correction is caused by the second-order correction, which was estimated to be of the order of 20%, leading to a maximum uncertainty of 1.5% in the corrected count rate. The total correction at the top of the strongest resonances and for dedicated beam pulses is always less than 6%. Although small, the dead-time effect is distorting the resonance shape and can hamper the resonance analysis. On the flat top of the 4.9-eV resonance, for example, this distortion is about 2%. Therefore, a corresponding correction was always applied. As a final remark, the structure between 12 and 14 MeV in the ETAC spectrum of Fig. 1is now explained by the pileup feature in Fig. 5. After the dead-time correction, a systematic check was carried out by comparing parasitic and dedicated pulses. Except for the 60.3-eV resonance, which shows the largest counting rate of >1µs−1with a corresponding dead time correction of 6%, the difference between the yields extracted for the two proton beam modes was found to be less than 1%, 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 58 60 62 En (eV) Capture yield Parasitic mode Dedicated mode 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 240 250 260 En (eV) Parasitic mode Dedicated mode FIG. 6. (Color online) Capture yields from data obtained in parasitic and dedicated beam mode for resonances with the largest dead-time corrections. as illustrated in Fig. 6, thus providing confidence in the validity of the dead-time correction. 2. Background The background components, which result from (i) in-beam γrays, (ii) ambient background, and (iii) αradioactivity of Ra contaminants in the scintillator [40] were studied by means of dedicated measurements and have been reduced by the conditions on ETAC discussed previously and by pulse-shape discrimination of the BaF2signals. Another background component is attributable to samplescattered neutrons. Scattered neutrons can be captured inside the TAC, mainly by the Ba isotopes of the scintillator, and may contaminate the capture yield from the Au sample. This background depends on the neutron sensitivity of the detector [41,42], which can be defined as the ratio between the efficiencies for detecting scattered neutrons, n, and capture events, γ. This background causes an artificial increase of the resonance area, particularly in resonances where the neutron width exceeds the radiative width. In Fig. 7the capture yield of the 197Au sample is compared with the yields from background runs with natC and natPb and without sample, which were all obtained with the analysis conditions described previously. The Pb measurement provides a good estimate of the sample-dependent component of the backgound in the Au measurement, being the areal density, the atomic number and the nonresonant elastic cross section of the two samples comparable. Because of the very similar shape of Pb and C yields, one can conclude that the background owing to in-beam γrays is small. Moreover, the Pb and C yields are close to the yield obtained without sample, demonstrating the low level of the residual background, which is attributable mainly to the neutron sensitivity of the detector. Several methods have been proposed to determine this background component [43,44]. In the present analysis the neutron sensitivity was determined from a measurement with a thick graphite sample. In this case, the measured count rates are 044616-6 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) 10-3 10-2 10-1 1 110 10210 3104 En (eV) Capture yield n + Au n + empty sample FIG. 7. (Color online) (Left) Experimental capture yield measured with Au and without sample. (Right) Experimental yields measured without sample and with the natCandnatPb samples, all analyzed with the same conditions for energy deposition and multiplicity. Resonances in the natPb yield are due to a small Sb contamination. weighted by the ratio of the capture and elastic cross sections of Au and C, respectively, n γ =CC CAu YAu γ YC n ,(4) where CCand CAu are the background corrected number of counts as a function of neutron energy. The capture yield for 197Au and the one for elastic scattering for 12C are calculated from the evaluated cross sections in the ENDF/B-VI library [45]. As shown in Fig. 8the neutron sensitivity of the 4πBaF2 detector is about 0.1, three orders of magnitude higher than that of the n TOF C6D6setup [27]. However, the background attributable to scattered neutrons can be reduced in an efficient way by suitable conditions on the total deposited energy ETAC . The optimal condition can be derived from the distributions of the total deposited energy shown in Fig. 1for Au, C, 10-2 10-1 1 10210 3104105 En (eV) Neutron sensitivity (εn/εγ) No conditions (i. e. Mγ≥2) Mγ≥3 ETAC>2.5 MeV, Mγ≥2 3.5<ETAC<7.5 MeV, Mγ≥2 FIG. 8. (Color online) The neutron sensitivity of the n TOF TAC as a function of neutron energy for different conditions. and an empty sample. Apart from the peak at 6.5 MeV related to 197Au(n,γ) reactions, structures at low energies are observed at 478 keV and 2.2 MeV associated with capture of scattered neutrons in the 10B capsules and in the hydrogen contained in the inner shell of absorbing material. The structure above 7.5 MeV is caused by capture of scattered neutrons by the odd Ba isotopes of the scintillator. The remaining background components falling within the selected window for the deposited energy are due to capture reactions on 19F and on the even Ba isotopes. However, these components are less evident in the ETAC spectrum of Fig. 1owing to the low capture cross sections and/or low natural abundances of these isotopes. The main contributions to the neutron sensitivity are summarized in Table II. The calculated neutron sensitivity is shown in Fig. 8for different conditions on ETAC and event multiplicity. The strongest reduction of about one order of magnitude is obtained with the condition used in the present analysis, that is, 3.5<E TAC <7.5 MeV and Mγ⩾2, by which events related to neutron capture in the 10B-loaded capsules, as well as in the neutron absorber, are rejected. The remaining background attributable to scattered neutrons is illustrated in Fig. 1by the difference between the spectra taken with the C sample and that of the empty sample position. For most resonances, this unavoidable background results in an increase of the neutron yield of a few percent, although it may reach up to 20% for resonances with very large n TABLE II. Main contributions to the neutron sensitivity of the TAC. Isotope Reaction Energy (MeV) Origin 10B(n,αγ) 0.48 Capsules 1H(n,γ) 2.2 n-absorber 138Ba (n,γ) 4.8 Scintillator 137Ba (n,γ) 8.6 Scintillator 135Ba (n,γ) 9.1 Scintillator 044616-7 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) 10-4 10-3 10-2 10-1 1 1 10 10210 3 En (eV) Capture yield n + Au neutron background FIG. 9. (Color online) Au capture yield and neutron background determined from experimental data using the second method described in the text. values. Therefore, a second step is needed in the analysis to subtract this residual background, either on the basis of resonance parameters from literature or directly from the present experimental data. In the first case, the neutron background Bn can be derived by scaling the measured carbon yield, Bn=(YC−Yempty)YAu n YC n ,(5) where YAu nand YC nare the elastic yields for Au and C from the evaluated ENDF/B-VI cross sections. The second method is described in more detail in Ref. [46] and relies exclusively on the measured ETAC spectra. In this case, the residual background in the Au sample is determined from the spectra measured with the Au and C sample (subtracted for the spectrum without sample), which are scaled to match the tail of the ETAC spectrum above 7.5 MeV (Fig. 9). This method relies on the assumption that only neutron capture on the odd Ba isotopes contributes above 7.5 MeV. Both techniques do have problems, however. The first method may suffer from the lack of reliable neutron widths in literature. It also tends to overestimate the background under the resonances, because multiple scattering in the absorber and in the BaF2scintillators distributes the corresponding background component over a larger TOF region beyond the resonance itself. Therefore, this method provides only an upper limit for the background. The second method may overestimate the background for the largest resonances because the ETAC region above 7.5 MeV can be affected by pileup events (see Figs. 5and 9). In the present analysis, the neutron background was determined by means of the second method using data obtained in parasitic mode, which are characterized by lower pileup. The resulting background was verified and confirmed by the results calculated via Eq. (5). An example of the residual neutron background is given in Fig. 9. This background and the associated statistical uncertainty are used as input in SAMMY code to analyze the experimental yields in terms of the R-matrix formalism. In this program the net capture yield Yγ(En)=Yexp(En)−Ybkg(En)(6) is determined by linear interpolation of the background. Therefore, the uncertainties of this background are directly reflected in the uncertainties of the resonance parameters from the TAC data given in Table III. IV. ANALYSIS OF THE C6D6DATA A. From measured count rate to capture yield The experimental yield is obtained as a function of neutron energy Enfrom the weighted count rate Nwas [25,47] Yexp =fNfexp Nw Ec ,(7) where Yexp,Nw, and depend on TOF or the neutron energy [for the sake of clarity, this dependence is omitted in Eq. (7)]. The weighting of the detector signals and the determination of Nware described in what follows. (En) is proportional to the total number of neutrons intersecting the sample with energy Enmeasured with the SiMon detector [36], and the flux calibration is contained in the yield normalization factor fN. The yield correction factor fexp accounts mostly for the 200-keV pulse-height threshold in the γ-ray detectors (Sec. IV C). Ecdenotes the total capture energy, which for a resonance at energy ERis given by Ec=Sn+ER, where Sn=6.512 MeV is the neutron separation energy of 197Au. The yield given by Eq. (7) still contains the contributions from capture and background events. The energy-dependent background level Ybkg(En) is determined from complementary measurements, as described in Sec. IV D. Finally, the net capture yield Yγ(En) is obtained by subtraction of the background as in Eq. (6). B. Pulse height weighting technique The pulse height weighting technique (PHWT) was introduced by Macklin and Gibbons [30] more than 40 years ago. By this technique the detector response is weighted in such a way that the detection efficiency becomes proportional to the energy of the registered γray. Before the first (n,γ) measurements at n TOF, there was no common recipe for determining the corresponding weighting factors (WFs) but different prescriptions have been used (see Refs. [48,49] for example). Furthermore, the WFs applied in the first decades of the PHWT had typical uncertainties of 20% [50]. Therefore, a substantial effort was dedicated at nTOF to validate the PHWT experimentally and to define a clear procedure to determine accurate WFs and the systematic uncertainty that can be achieved with this technique. It was shown that an uncertainty of better than 2% can be ascribed to the PHWT provided that realistic MC simulations of the experimental setup and of the capture events are included in the analysis [25]. At present, there is general agreement that the MC approach represents the correct method for the calculation of the WFs [25,42,51,52]. 044616-8 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) TABLE III. Resonance parameters extracted from the R-matrix analysis of the n TOF C6D6and TAC data. Quantum numbers, Jand , taken form ENDF/B-VII. Parameters in square brackets have to be taken with caution (see text). Parameters in parentheses indicate cases with large systematic uncertainties. ER(eV) J C6D6TAC n(meV) γ(meV) gnγ/ (meV) n(meV) γ(meV) gnγ/ (meV) 4.905 2 0 15.2 124 8.5 15.2 124 8.5 (36.07)a 46.63 1 0 0.223 ±0.006 128 0.084 ±0.002 0.220 ±0.008 128 0.082 ±0.003 58.02 1 0 4.34 ±0.05 112 1.57 ±0.02 4.43 ±0.05 112 1.60 ±0.02 60.23 2 0 [73.93 ±0.29] 110 27.63 ±0.12 [72.7±0.4] 110 27.36 ±0.09 78.44 1 0 17.79 ±0.18 120 5.81 ±0.06 16.6±0.2 120 5.48 ±0.05 107.0 2 0 8.29 ±0.10 110 4.82 ±0.06 7.9±0.1 110 4.63 ±0.05 122.2 2 0 0.86 ±0.02 128 0.53 ±0.01 0.89 ±0.03 128 0.55 ±0.02 144.3 1 0 9.35 ±0.15 120 3.25 ±0.05 8.8±0.2 120 3.08 ±0.05 151.3 2 0 [21.6±0.4] [149.7 ±5.1] 11.8±0.6[22.7±0.4] [141 ±5] 12.2±0.2 162.9 1 0 [42.5±1.1] [196.8 ±8.0] 13.1±0.8[46±1] [170 ±7] 13.6±0.3 165.0 2 0 8.67 ±0.14 109 5.02 ±0.08 9.1±0.2 109 5.24 ±0.09 189.9 1 0 [49.8±0.9] 130 13.5±0.2[48.13 ±0.9] 130 13.2±0.2 209.0b10 [0.86 ±0.04] [181.9 ±28.1] 0.32 ±0.07 [0.87 ±0.09] [190 ±60] 0.32 ±0.03 240.4 2 0 [86.6±2.1] [99.6 ±1.9] 29.0±1.0[82±7] [98 ±7] 27.9±1.6 255.4b10 [0.50 ±0.05] [129.2 ±39.7] 0.19 ±0.08 [0.58 ±0.09] [120 ±60] 0.22 ±0.03 262.1 1 0 [151.7±3.7] [124.0 ±2.4] 25.6±0.9 [167 ±8] [108 ±3] 24.6±0.6 273.7 2 0 4.41 ±0.12 110 2.65 ±0.07 5.0±0.2 110 3.0±0.1 293.2 2 0 [347.7±4.6] [123.6 ±1.5] 57.0±1.2 [336 ±7] [128 ±2] 57.9±0.7 329.2 2 0 [41.5±1.0] 137 19.9±0.5[42±1] 137 20.2±0.4 330.6 1 0 [56.2±1.8] 130 14.7±0.5[59±2] 130 15.3±0.4 355.3 2 0 37.6±0.9 125 18.1±0.437.8±1.0 125 18.13 ±0.4 370.9 2 0 [108.9±3.6] 99 32.4±1.2 [101 ±4] 99 31.3±0.6 375.4 1 0 12.32 ±0.42 125 4.21 ±0.14 12.5±0.6 125 4.3±0.2 381.8 2 0 [73.9±2.4] 97 26.2±0.9[70±2] 97 25.5±0.5 400.1 2 0 6.08 ±0.23 128 3.63 ±0.14 6.4±0.4 128 3.8±0.2 401.3 1 0 25.8±0.8 140 8.2±0.325±1 140 7.9±0.3 440.1 1 0 281.4 [149.3 ±3.0] 36.6±0.8 281.4 [129 ±3] 33.1±0.6 450.8 2 0 [63.7±2.0] 110 25.2±0.8[67±2] 110 26.0±0.6 477.1 2 0 296.1 [124.9 ±2.2] 54.9±1.0 296.1 [118 ±3] 52.8±0.8 489.5 1 0 [62.2±2.1] 138 16.1±0.6[57±2] 138 15.1±0.4 493.6 2 0 28.5±0.8 111 14.2±0.426.4±1.0 111 13.3±0.4 533.6 2 0 31.8±1.0 130 16.0±0.532.5±0.5 130 16.2±0.2 548.1 1 0 [58.6±2.2] 127 15.0±0.6[61±1] 127 15.4±0.2 561.2 2 0 2.44 ±0.19 128 1.49 ±0.12 2.5±0.1 128 1.52 ±0.07 578.5 2 0 288.4 [126.0 ±3.1] 54.8±1.4 288.4 [132 ±2] 56.7±0.5 580.4 1 0 306.8 [121.7 ±4.0] 32.7±1.1 306.8 [103 ±2] 28.9±0.4 586.3 2 0 22.4±0.8 134 12.0±0.422 .4±0.4 134 12.0±0.2 602.4 2 0 223.9 [112.6 ±2.6] 46.8±1.1 223.9 [113 ±1] 47.0±0.4 616.9 1 0 [111.1±5.2] 135 22.9±1.2 [117 ±3] 135 23.5±0.3 624.1 1 0 [53.4±2.3] 121 13.9±0.6[53±1] 121 13.9±0.2 627.9 2 0 24.7±0.8 138 13.1±0.525.3±0.4 138 13.4±0.2 638.3 2 0 464.0 [118.5 ±2.5] 59.0±1.3 464.0 [118 ±1] 58.9±0.5 658.4 2 0 4.21 ±0.30 97 2.52 ±0.18 4.7±0.297 2.8±0.1 685.6 1 0 16.8±0.9 128 5.6±0.318.7±0.5 128 6.12 ±0.15 695.3 1 0 666.7 138.5 ±4.0 43.0±1.3 666.7 128 ±240.3±0.5 698.5 2 0 736.1 109.8 ±3.2 59.7±1.8 736.1 115 ±162.2±0.6 715.2 2 0 [111.0±20.2] [112.0 ±18.0] 34.8±9.4 [105.7±10.0] [120 ±5] 35 ±2 738.0 1 0 10.6±0.7 120 3.7±0.211.3±0.4 120 3.9±0.14 759.5 1 0 426.7 [110.3 ±3.2] 32.9±1.0 426.7 116 ±234.2±0.4 773.3 1 0 474.6 [125.0 ±3.8] 37.1±1.1 474.6 [127 ±2] 37.6±0.4 783.8 2 0 [14.0±4.9] 140 37.3±1.9 [102 ±2] 140 36.8±0.4 795.5 2 0 177.6 [122.7 ±4.2] 45.4±1.7 177.6 [124 ±2] 45.7±0.4 812.8 1 0 22.1±1.3 128 7.1±0.423.4±0.7 128 7.4±0.2 044616-9 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) (eV) n E 1 10 2 10 3 10 4 10 5 10 Weighted Counts (arb. units) 1 10 2 10 3 10 4 10 Normalization of Background Components Gold+Filters -raysγIn-beam Neutron Activation Bkg Total Normalized Bkg FIG. 14. (Color online) Capture yield of Au with Al and W filters in the beam to determine the background level. D. Background Apart from the effects discussed so far, there are two remaining sources of background at the measuring station. The component between thermal and ∼200 eV (see Fig. 13)arises from neutrons scattered in the sample, which are thermalized and captured somewhere in the experimental area. This ambient background level is rather small. Contributions from neutrons that do not interact with the sample are practically negligible; thus, the neutron beam line continues about 12 m beyond the capture experiment to the beam dump, which is separated from the experimental area by a thick concrete wall. The second type of background contributes significantly in the energy range between 200 eV and ∼500 keV (see Fig. 13) and it is due to delayed in-beam γrays from neutron captures in the water moderator of the spallation source. In fact, these γrays can be scattered by the sample. The origin and the energy dependence of both background components was determined from a dedicated measurement with a lead sample as shown in Fig. 13. The relative contribution from each background component to the measurement of the gold sample was obtained in a series of measurements with the gold sample in combination with neutron filters of Al and W. The filters, which are installed at a flight path of 135 m, were thick enough that the strongest resonances became black, thus showing the effective background level at these neutron energies. The effect of the black resonances on the Au capture yield was fitted as a smooth perturbation of the reaction yield, as illustrated in Fig. 14. V. R E S U LT S A. From capture yield to cross section The self-shielding corrected capture yield Yth =(1 −e−nσtot )σγ σtot (15) is related to the capture cross section σγand to the total cross section σtot, where nis the areal density of the sample. The net capture yields in the RRR obtained from the TAC and C6D6data can be expressed in terms of the resonance parameters, Yγ(En)=Yγ(En,a),(16) where the vector of resonance parameters a=Jπ i,,E R,i, γ,i, n,i with i=1,m (17) includes the m=268 resonances contained in both data sets in the energy range between 1 eV and 5 keV. (eV) n E 45.5 46 46.5 47 47.5 48 Capture Yield 0.01 0.02 0.03 0.04 0.05 (eV) n E 330 340 350 360 370 380 390 400 Capture Yield 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 (eV) n E 440 450 460 470 480 490 Capture Yield 0.05 0.1 0.15 0.2 0.25 0.3 0.35 (eV) n E 955 960 965 970 975 980 985 990 995 Capture Yield 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 FIG. 15. (Color online) Examples for resonance fits of the C6D6data (solid symbols with error bars). The blue curve represents the evaluated data listed in ENDF/B-VII database [16] and the red curve shows the present fits. 044616-16 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 45.5 46 46.5 47 47.5 48 En (eV) Capture yield 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 950 960 970 980 990 1000 En (eV) Capture yield FIG. 16. (Color online) Examples for resonance fits of the TAC data (solid symbols with error bars). The blue curve represents the evaluated data listed in ENDF/B-VII database [16] and the red curve shows the present fits. These capture yields were analyzed using the R-matrix analysis code SAMMY [18]. The yield is parametrized via the multilevel Breit Wigner (MLBW) formalism using a scattering radius of 9.658 fm [17] and a temperature of 293 K for the correction of the Doppler effect. Other experimental effects, that is, multiple neutron scattering in the sample and neutron self-shielding, are properly taken into account within the SAMMY code. Although for the majority of the resonances the width is dominated by the Doppler broadening, the resonance broadening due to the n TOF resolution function is also considered in the SAMMY fits by the implemented RPI parameterization [18]. The parameters for this function were determined in a series of measurements on narrow resonances at higher neutron energies [24]. The fitting procedure allowed us to extract the resonance parameters from the measured capture yields, although in many cases only the resonance energy ERand the total capture kernel gγn/ should be considered as the real measurable quantities. Only in cases where one of the channels dominates over the other, nγor nγ, the resonance shape analysis is sufficiently sensitive to the smaller value. Otherwise, the resonance parameters are given in square brackets in Table III to indicate that they have to be taken with some caution. The quoted resonance energies are the ones measured with the C6D6detectors. Resonance parameters from the ENDF/B-VII data [16] library were used as initial values in our fits. In general, we tried to vary as few parameters as possible (either nor γ), but when the improvement in the χ2value of the fit was substantial, both parameters (nand γ) were allowed to vary. Because both analyses were performed independently, the comparison of the resulting data sets provides an estimate of the overall systematic uncertainty and reveals the benefits and drawbacks of both capture detectors. However, a common R-matrix analysis of both data sets is beyond the scope of the present work. In any case, a complete re-evaluation of the cross section and its potential recognition as a capture standard requires also the combined analysis of data sets, including transmission data, of other facilities. In this context, the present results provide relevant information on the capture kernels and in some cases on the resonance parameters. Examples of the fitted yield are shown in Figs. 15 and 16, where the data points stand, respectively, for the C6D6and TAC measurement, the dotted curve represents the evaluated data listed in ENDF/B-VII database [16], and the solid curve the actual MLBW fits to the present data. The resonance parameters and the radiative kernels derived from the C6D6and TAC data are listed in Table III. Parameters without uncertainties have been kept fixed in the analysis. Otherwise, the standard deviation of the fitted parameter is quoted as the corresponding uncertainty. B. Discussion of uncertainties The total uncertainties in the capture yields measured with the TAC and the C6D6detectors are summarized in Table V. Apart from the contribution by the energy dependence of the neutron flux, the total uncertainties of the two methods are dominated by completely independent components. In both cases, the largest uncertainty of 3% is caused by the energy dependence of the neutron flux. TABLE V. Different components of estimated systematic or correlated uncertainty in the measured capture yields. Component Uncertainty (%) C6D6TAC PHWT 2 – Normalization 0.51 Dead time – 0–1.2 Background 1 1 Flux shape 3 3 Beam profile [62]0.62 Total 3.8 3.9–4.1 044616-17 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) 0.5 1 1.5 2 2.5 3 3.5 0 1000 2000 3000 4000 5000 ER (eV) Kernel ratio (TAC/C6D6) 0 5 10 15 20 25 30 0.6 0.8 1 1.2 1.4 1.6 0 5 10 15 20 25 30 0.6 0.8 1 1.2 1.4 1.6 Kernel ratio (TAC/C6D6) Number of cases FIT (χ2=1.73): Constant 16.88 Mean 1.015 Sigma 0.08 FIG. 17. (Left) Ratio of capture kernels versus resonance energy. (Right) The statistical distributions of the kernel ratios agree within 1.5% at an average standard deviation of about 8%. A more specific discussion of the TAC-related uncertainties shows that a 2% uncertainty is caused by the fact that the sample was smaller than the diameter of the neutron beam. The difference between the related normalization factors extracted from parasitic and dedicated data is about 0.5%. From this difference and from the uncertainty of the dead-time correction in the 4.9-eV resonance, an uncertainty of 1% was estimated for the normalization by the saturated resonance technique. The dead-time correction as such was significant only for the stronger resonances. The background determined by the measurement without sample was normalized by means of the number of protons per pulse, which carries an uncertainty of 1%. The uncertainty due to the neutron sensitivity of the TAC is taken into account in the resonance shape analysis. The situation for the C6D6data differs in several respects. In this case the 2% uncertainty of the PHWT contributed strongly to the overall uncertainty. While the uncertainty of the background subtraction was similar to the TAC analysis, the normalization procedure was not affected by dead-time effects, which were negligible for the smaller C6D6detectors. Moreover, the influence of the detector thresholds is small because the γ-ray spectra for the different resonances are quite similar because of the large number of levels available for decay after neutron capture. Therefore, the normalization of the C6D6data is affected by an uncertainty of only 0.5%. The sum of these components yields an overall systematic uncertainty of 3.8% for the C6D6and 3.9–4.1% (depending on the count rate) for the TAC results, as listed in Table V. C. Comparison between TAC and C6D6results The capture kernel gnγ/ is a quantity proportional to the area of a resonance, which is sensitive to systematic effects related to the measurement technique. Therefore, the ratio of the capture kernels obtained from the TAC and C6D6data is discussed with respect to several quantities and their respective systematic uncertainties. The ratio in Fig. 17 shows good agreement between the TAC and C6D6results with an average deviation of less than 2%, thus confirming that the absolute normalization of the capture yield in both independent analyses was consistent and reliable. As shown in Fig. 17 (right), the standard deviation between both data sets is 8%, as expected due to different experimental effects in the two data sets as well as to the uncertainty introduced by the resonance shape analysis. 0.5 1 1.5 2 2.5 3 3.5 0 50 100 (a) Kernel (meV) Kernel ratio (TAC/C6D6) 10 -1 1 Count rate (µs-1) (b) FIG. 18. Comparison of the capture kernels obtained from the TAC data and C6D6data as a function of resonance area (a) and peak resonance count rate of TAC data (b). 044616-18 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) 0.5 1 1.5 2 2.5 3 3.5 10 -1 1 10 Γn/Γγ Kernel ratio (TAC/C6D6) (a) 1 10 10210 3 gΓn (meV) (b) FIG. 19. Comparison of the capture kernels obtained from the TAC and C6D6data versus n/ γratio (a) and the resonance strength gn(b). The absence of a correlation between the kernels and their ratios provides evidence for the accuracy of WF as illustrated in Fig. 18(a) and indicates that the γ-ray attenuation in the sample was properly treated in the C6D6data. The figure shows also that larger deviations are related to weak resonances. The overall good agreement indicates also that the individual resonance-related corrections applied in the analysis of the TAC data, namely, neutron sensitivity and dead-time corrections, are consistent with the C6D6results, where these corrections have practically no impact. Figure 18(b) illustrates that the kernel ratio as a function of TAC count rate is fully consistent for the strong resonances (and high counting rates). Only for small resonances there are discrepancies between the two data sets, similar to the case of Fig. 18(a). The proper treatment of the neutron-sensitivity correction in the TAC data is supported by the absence of a clear trend between the kernel ratios and the n/ γratios shown in Fig. 19(a). This main correction for the TAC data is further confirmed by the plot of the kernel ratio versus resonance strength gnin Fig. 19(b). 10-4 10-3 10-2 10-1 1 200 250 300 En (eV) Capture yield ENDF/B-VII TAC data 0 0.01 4.9 5 En (keV) ENDF/B-VII TAC data FIG. 20. (Color online) Capture yield reconstructed from the ENDF/B-VII parameters and the corresponding TAC data. (a) The resonances at 209 and 255 eV are not included in the evaluation. (b) The evaluation ends at 4.83 keV. D. Comparison with libraries With respect to the ENDF/B-VII evaluation, we observed: (i) three resonances not included in the evaluation, at 209, 255, and 1634 eV (Fig. 20,left); (ii) six new resonances at 4869.1, 4880.4, 4892.2, 4915.7, 4944.0, and 5000.4 eV in the energy range above previous analyses (Fig. 20, right); (iii) a structure at 36.07 eV, that is probably a resonance; (iv) resonances at 2286.4, 3214.8, and 4315.4 eV, which are probably doublets; (v) features at 2652.6, 2896.1, 3302.3, 3310.0, and 4046.7 eV, which are probably due to multiple scattering in nearby resonances rather than genuine resonances as quoted in ENDF/B-VII (Fig. 21). On average, the present capture kernels are 10–15% lower than the values from ENDF/B-VII. There is a systematic trend for the discrepancies to increase with neutron energy as shown 10-6 10-5 10-4 10-3 10-2 10-1 2.6 2.62 2.64 2.66 2.68 2.7 En (keV) Yield from ENDF parameters Unbroadened yield multiple scattering FIG. 21. (Color online) Capture yield calculated from the ENDF/B-VII parameters (dotted line) and folded with the experimental resolution (thick solid line) compared to the contribution from multiple scattering, which was calculated without the resonance at 2652.6 (dotted line). The structure indicated by the arrow is probably due to multiple scattering. 044616-19 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) 0 5 10 15 20 25 0.5 1 1.5 2 2.5 ID Entries Mean RMS 100 254 1.156 0.3191 Kernel ratio (ENDF/C6D6) Number of cases FIG. 22. (Left) Ratio of capture kernels from ENDF/B-VII and the present C6D6results vs. resonance energy. (Right) The corresponding distribution of the ratios. for the C6D6and the TAC data in the left panels of Figs. 22 and 23, respectively. VI. CONCLUSIONS The 197Au(n,γ) reaction has been measured at n TOF with the aim of improving the accuracy of the neutron capture cross section in the RRR. To identify and minimize systematic uncertainties, especially those related to the detection efficiency and to the neutron sensitivity, two conceptually different detection systems have been employed, a TAC, and a total energy system based on hydrogen-free C6D6liquid scintillators. The conditions used in the analysis of the TAC data were suitably chosen to reduce backgrounds as far as possible, notably the effect of neutron scattering from the sample. Corrections were applied to account for the dead time and the neutron sensitivity of the TAC data. For the C6D6measurement the accurate weighting function technique was used, which was developed for the n TOF setup. The data were then corrected for threshold effects and for electron conversion of the capture γrays in the sample. The resonances measured with the two different systems were separately analyzed with the code SAMMY to extract the radiative kernels. A total of 268 resonances, from 1 eV to about 5 keV were analyzed. The comparison of the results obtained with the two detectors showed very good agreement with an average systematic difference of less than 2%, thus confirming the high accuracy of the data. A standard deviation of 8% between the capture kernels extracted from the two data sets indicates that an uncertainty of less than 5% per detector has been reached for most of the resonances. Some larger differences may partially be attributed to residual effects related to the neutron sensitivity of the TAC or to the background subtraction for the smallest resonances. Six new resonances have been identified in the energy region between 4.83 and 5 keV. Three resonances that had been removed 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 1000 2000 3000 4000 5000 ER (eV) Kernel ratio (ENDF/TAC) 0 5 10 15 20 25 0.5 1 1.5 2 2.5 ID Entries Mean RMS 100 254 1.145 0.3003 Kernel ratio (ENDF/TAC) Number of cases FIG. 23. (Left) Ratio of capture kernels from ENDF/B-VII and the present TAC results vs resonance energy. (Right) The corresponding distribution of the ratios. 044616-20 197Au(n,γ) CROSS SECTION IN THE ... PHYSICAL REVIEW C 81, 044616 (2010) in the ENDF/B-VII evaluation compared to the ENDF/B-VI version have been clearly observed. On the contrary, five resonances listed in the evaluations seem to be caused by multiple scattering background from nearby resonances. A combination of the two data sets will make it possible to obtain results with an estimated uncertainty close to 4% for the capture resonances. In combination with high-accuracy transmission data, the present results may be used to extract very accurate resonance parameters, eventually leading to the extension of the Au cross section as a capture standard in the RRR. ACKNOWLEDGMENTS This work was supported by the European Commission’s 5th Framework Programme under Contract No. FIKW-CT- 2000-00107 (n TOF-ND-ADS Project) and by the funding agencies of the participating institutes. [1] G. Wallerstein et al.,Rev. Mod. Phys. 69, 995 (1997). [2] F. K¨ appeler, Prog. Part. Nucl. Phys. 43, 419 (1999). [3] W. Gudowski, Nucl. Phys. A 752, 623 (2005). [4] NEA-OEDC Report, 2002. [5] M. Salvatores and I. Slessarev, and V. Berthou, Prog. Nucl. Energy 38, 167 (2001). [6] A. D. Carlson, V. G. Pronyaev, D. L. Smith, N. M. Larson, Zhenpeng Chen, G. M. Hale, F.-J. Hambsch, E. V. Gai, Soo-Youl Oh, S. A. Badikov, T. Kawano, H. M. Hofmann, H. Vonach, and S. Tagesen, Nucl. Data Sheets 110, 3215 (2009). [7] L. R. Macklin, J. Halperin, and R. R. Winters, Nucl. Instrum. Methods 164, 213 (1979). [8] J. S. Desjardins, J. L. Rosen, W. W. Havens, and J. Rainwater, Phys. Rev. 120, 2214 (1960). [9] M. Asghar, M. C. Moxon, and C. M. Chaffey, in Proceedings of the International Conference on the Study of Nuclear Structure with Neutrons,editedbyM.N´ eve de M´ evergines, P. Van Assche, and J. Verneir (North–Holland, Amsterdam, 1965), p. 522. [10] J. Julien, S. D. Barros, G. Bianchi, C. Corge, V. D. Huynh, G. L. Poittevin, J. Morgenstern, F. Netter, C. Samour, and R. Vastel, Nucl. Phys. 76, 391 (1966). [11] J. Lynn and E. Rae, J. Nucl. Energy 4, 418 (1957). [12] R. N. Alves, J. Julien, J. Morgenstern, and C. Samour, Nucl. Phys. A 131, 450 (1969). [13] R. L. Macklin, J. Halperin, and R. R. Winters, Phys. Rev. C 11, 1270 (1975). [14] M. M. Hoffman, W. M. Sanders, G. J. Berzins, L. J. Brown, and D. D. Phillips, Los Alamos National Lab Report No. LA-5552, 1974. [15] P. G. Young, Los Alamos National Lab, USA, ENDF/B-VI Evaluation, MAT # 7925, 1984, revision 1, July 1991, data retrieved from the ENDF database. [16] P. G. Young, Los Alamos National Lab, USA, ENDF/B-VI Evaluation, MAT # 7925, 1984, revision 2, Dec. 2006, data retrieved from the ENDF database. [17] S. F. Mughabghab, Neutron Cross Sections: Neutron Resonance Parameters and Thermal Cross Sections (Academic Press, New York, 2006). [18] N. M. Larson, Oak Ridge National Laboratory Report No. ORNL/TM-9179/R7. [19] C. Lederer et al. (unpublished). [20] S. Abramovich et al., CERN Technical Report No. CERNSPSC-99-8, SPSC-P-310, 1999. [21] U. Abbondanno et al., CERN Technical Report CERN-SL-2002- 053 ECT, 2003. [22] F. Gunsing et al. (The n TOF Collaboration), Nucl. Instrum. Methods B 261, 925 (2007). [23] J. Pancin et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 524, 102 (2004). [24] G. Lorusso et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 532, 622 (2004). [25] U. Abbondanno et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 521, 454 (2004). [26] Acqiris SA, 18, chemin des Aulx, CH-1228 Plan-les-Ouates, Switzerland, at present information can be found in Agilent Technologies, http://www.agilent.com. [27] R. Plag, M. Heil, F. K¨ appeler, P. Pavlopoulos, R. Reifarth, and K. Wisshak, Nucl. Instrum. Methods A 496, 425 (2003). [28] K. H. Guber, L. C. Leal, R. O. Sayer, P. E. Koehler, T. E. Valentine, H. Derrien, and J. A. Harvey, Nucl. Instrum. Methods B241, 218 (2005). [29] K. H. Guber, L. C. Leal, R. O. Sayer, P. E. Koehler, T. E. Valentine, H. Derrien, and J. A. Harvey, AIP Conf. Ser. 769, 1706 (2005). [30] R. L. Macklin and J. H. Gibbons, Phys. Rev. 159, 1007 (1967). [31] C. Guerrero et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 608, 424 (2009). [32] E. Berthoumieux et al. (The n TOF Collaboration), in Proceedings of the International Conference on Nuclear Data for Science and Technology, Nice, France, 2007, edited by O. Bersillon et al., EDP Sciences (CEA, Paris, 2008, 2007), p. 627. [33] C. Massimi et al. (The n TOF Collaboration), in Proceedings of the International Conference on Nuclear Data for Science and Technology, Nice, France, 2007, edited by O. Bersillon et al., EDP Sciences (CEA, Paris, 2008, 2007), p. 1265. [34] C. Lampoudis et al. (The n TOF Collaboration), in Proceedings of the International Conference on Nuclear Data for Science and Technology, Nice, France, 2007, edited by O. Bersillon et al., EDP Sciences (CEA, Paris, 2008, 2007), p. 59. [35] C. Guerrero et al. (The n TOF Collaboration), in Proceedings of the International Conference on Nuclear Data for Science and Technology, Nice, France, 2007, edited by O. Bersillon et al., EDP Sciences (CEA, Paris, 2008, 2007), p. 627. [36] S. Marrone et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 517, 389 (2004). [37] C. Borcea et al.,Nucl. Instrum. Methods A 513, 524 (2003). [38] S. Marrone et al.,Nucl. Instrum. Methods A 568, 904 (2006). [39] G. F. Knoll, Radiation Detection and Measurement (Wiley, New York, 1989). [40] K. Wisshak, K. Guber, F. K¨ appeler, J. Krisch, H. M¨ uller, G. Rupp, and F. Voss, Nucl. Instrum. Methods A 292, 595 (1990). [41] F. Corvi, Nuclear Energy Agency Report No. NEA/NSC/ DOC(95)1, 1994. 044616-21 C. MASSIMI et al. PHYSICAL REVIEW C 81, 044616 (2010) [42] A. Borella, G. Aerts, F. Gunsing, M. Moxon, P. Schillebeeckx, and R. Wynants, Nucl. Instrum. Methods A 577, 626 (2007). [43] M. Heil, R. Reifarth, M. M. Fowler, R. C. Haight, F. K¨ appeler, R. S. Rundberg, E. H. Seabury, J. L. Ullmann, J. B. Wilhelmy, and K. Wisshak, Nucl. Instrum. Methods A 459, 229 (2001). [44] R. Reifarth, T. A. Bredeweg, A. Alpizar-Vicente, J. C. Browne, E. I. Esch, U. Greife, R. C. Haight, R. Hatarik, A. Kronenberg, J. M. O’Donnell et al.,Nucl. Instrum. Methods A 531, 530 (2004). [45] M. B. Chadwick and P. G. Young, Los Alamos National Lab, USA, ENDF/B-VI Evaluation, MAT # 600, 1996, revision 2, June 1996, data retrieved from the ENDF database. [46] K. Wisshak et al.,Phys. Rev. C 73, 045807 (2006). [47] C. Domingo-Pardo et al. (The n TOF Collaboration), Phys. Rev. C74, 025807 (2006). [48] F. Corvi, A. Prevignano, H. Liskien, and P. B. Smith, Nucl. Instrum. Methods A 265, 475 (1988). [49] F. G. Perey, J. O. Johnson, T. A. Gabriel, R. L. Macklin, R. R. Winters, J. H. Todd, and N. H. Hill, in Proceedings of the International Conference on Nuclear Data for Science and Technology (Mito, Japan, 1988), pp. 379–382. [50] R. L. Macklin, Nucl. Sci. Eng. 95, 200 (1987). [51] P. E. Koehler, R. R. Spencer, R. R. Winters, K. H. Guber, J. A. Harvey, N. W. Hill, and M. S. Smith, Phys.Rev.C54, 1463 (1996). [52] J. N. Wilson, B. Haas, S. Boyer, D. Dassie, G. Barreau, M. Aiche, S. Czajkowski, C. Grosjean, and A. Guiral, Nucl. Instrum. Methods A 511, 388 (2003). [53] S. Agostinelli et al. (Geant4 Collaboration), Nucl. Instrum. Methods A 506, 250 (2003). [54] C. Domingo-Pardo, Ph.D. thesis, CSIC-University of Valencia, 2004. [55] C. Domingo-Pardo et al. (The n TOF Collaboration), Phys. Rev. C74, 055802 (2006). [56] C. Domingo-Pardo et al. (The n TOF Collaboration), Phys. Rev. C75, 015806 (2007). [57] C. Domingo-Pardo et al. (The n TOF Collaboration), Phys. Rev. C76, 045805 (2007). [58] J. L. Tain and D. Cano-Ott, Nucl. Instrum. Methods A 571, 719 (2007). [59] J. L. Tain and D. Cano-Ott, Nucl. Instrum. Methods A 571, 728 (2007). [60] C. Guerrero, D. Cano-Ott, E. Mendoza, and J. L. Ta´ ın, AIP Conf. Ser. 1090, 372 (2009). [61] U. Abbondanno et al. (The n TOF Collaboration), Nucl. Instrum. Methods A 538, 692 (2005). [62] S. Marrone et al.,Phys.Rev.C73, 034604 (2006). 044616-22