Performance of the neutron time-of-flight facility n_TOF at CERN

DOI 10.1140/epja/i2013-13027-6 Tools for Experiment and Theory – Experimental Physics Eur. Phys. J. A (2013) 49:27 THE EUROPEAN PHYSICAL JOURNAL A Performance of the neutron time-of-flight facility n TOF at CERN C. Guerrero1,a, A. Tsinganis1,2, E. Berthoumieux3,1, M. Barbagallo4, F. Belloni3, F. Gunsing3, C. Weiß5,1, E. Chiaveri3,1, M. Calviani1, V. Vlachoudis1, S. Altstadt6, S. Andriamonje1, J. Andrzejewski7, L. Audouin8, V. B´ecares9,F.Beˇcv´aˇr10,J.Billowes 11, V. Boccone1,D.Bosnar 12,M.Brugger 1, F. Calvi˜no13, D. Cano-Ott9, C. Carrapi¸co14,F.Cerutti 1, M. Chin1, N. Colonna4, G. Cort´es13, M.A. Cort´es-Giraldo15, M. Diakaki2, C. Domingo- Pardo16,I.Duran 17, R. Dressler18,N.Dzysiuk 19, C. Eleftheriadis20, A. Ferrari1,K.Fraval 3,S.Ganesan 21, A.R. Garc´ıa9,G.Giubrone 16,K.G¨obel6,M.B.G´omez-Hornillos13, I.F. Gon¸calves14,E.Gonz´alez-Romero9, E. Griesmayer5, P. Gurusamy21, A. Hern´andez-Prieto1,13, P. Gurusamy21, D.G. Jenkins22,E.Jericha 5,Y.Kadi 1, F. K¨appeler23, D. Karadimos2,N.Kivel 18, P. Koehler24,M.Kokkoris 2,M.Krtiˇcka10,J.Kroll 10, C. Lampoudis3, C. Langer6, E. Leal-Cidoncha17, C. Lederer6,25, H. Leeb5, L.S. Leong8, R. Losito1, A. Manousos20, J. Marganiec7, T. Mart´ınez9, C. Massimi26, P.F. Mastinu19, M. Mastromarco4, M. Meaze4, E. Mendoza9, A. Mengoni27, P.M. Milazzo28, F. Mingrone26,M.Mirea 29, W. Mondalaers30, T. Papaevangelou3, C. Paradela17,A.Pavlik 25, J. Perkowski7, A. Plompen30, J. Praena15, J.M. Quesada15, T. Rauscher31,R.Reifarth 6,A.Riego 13, F. Roman1,29, C. Rubbia1,32, M. Sabate-Gilarte15,R.Sarmento 14, A. Saxena21, P. Schillebeeckx30, S. Schmidt6,D.Schumann 18, P. Steinegger18, G. Tagliente4, J.L. Tain16,D.Tarr´ıo17,L.Tassan-Got 8,S.Valenta 10, G. Vannini26, V. Variale4, P. Vaz14,A.Ventura 27,R.Versaci 1, M.J. Vermeulen22, R. Vlastou2,A.Wallner 25,T.Ware 11, M. Weigand6, T. Wright11,andP.ˇ Zugec12 1European Organization for Nuclear Research (CERN), Geneva, Switzerland 2National Technical University of Athens (NTUA), Greece 3Commissariat `al’ ´ Energie Atomique (CEA) Saclay - Irfu, Gif-sur-Yvette, France 4Istituto Nazionale di Fisica Nucleare, Bari, Italy 5Atominstitut, Technische Universit¨at Wien, Austria 6Johann-Wolfgang-Goethe Universit¨at, Frankfurt, Germany 7Uniwersytet L´odzki, Lodz, Poland 8Centre National de la Recherche Scientifique/IN2P3 - IPN, Orsay, France 9Centro de Investigaciones Energeticas Medioambientales y Tecnol´ogicas (CIEMAT), Madrid, Spain 10 Charles University, Prague, Czech Republic 11 University of Manchester, Oxford Road, Manchester, UK 12 Department of Physics, Faculty of Science, University of Zagreb, Croatia 13 Universitat Politecnica de Catalunya, Barcelona, Spain 14 Instituto Tecnol´ogico e Nuclear, Instituto Superior T´ecnico, Universidade T´ecnica de Lisboa, Lisboa, Portugal 15 Universidad de Sevilla, Spain 16 Instituto de F´ısica Corpuscular, CSIC - Universidad de Valencia, Spain 17 Universidade de Santiago de Compostela, Spain 18 Paul Scherrer Institut, Villigen PSI, Switzerland 19 Istituto Nazionale di Fisica Nucleare, Laboratori Nazionali di Legnaro, Italy 20 Aristotle University of Thessaloniki, Thessaloniki, Greece 21 Bhabha Atomic Research Centre (BARC), Mumbai, India 22 University of York, Heslington, York, UK 23 Karlsruhe Institute of Technology, Campus Nord, Institut f¨ur Kernphysik, Karlsruhe, Germany 24 Oak Ridge National Laboratory (ORNL), Oak Ridge, TN 37831, USA 25 University of Vienna, Faculty of Physics, Austria 26 Dipartimento di Fisica, Universit`a di Bologna, and Sezione INFN di Bologna, Italy 27 Agenzia nazionale per le nuove tecnologie, l’energia e lo sviluppo economico sostenibile (ENEA), Bologna, Italy 28 Istituto Nazionale di Fisica Nucleare, Trieste, Italy 29 Horia Hulubei National Institute of Physics and Nuclear Engineering - IFIN HH, Bucharest - Magurele, Romania 30 European Commission JRC, Institute for Reference Materials and Measurements, Retieseweg 111, B-2440 Geel, Belgium 31 Department of Physics and Astronomy - University of Basel, Basel, Switzerland 32 Laboratori Nazionali del Gran Sasso dell’INFN, Assergi (AQ), Italy Page 2 of 15 Eur. Phys. J. A (2013) 49:27 Received: 23 November 2012 / Revised: 4 February 2013 Published online: 25 February 2013 c The Author(s) 2013. This article is published with open access at Springerlink.com Communicated by D. Pierroutsakou Abstract. The neutron time-of-flight facility n TOF features a white neutron source produced by spallation through 20 GeV/cprotons impinging on a lead target. The facility, aiming primarily at the measurement of neutron-induced reaction cross sections, was operating at CERN between 2001 and 2004, and then underwent a major upgrade in 2008. This paper presents in detail all the characteristics of the new neutron beam in the currently available configurations, which correspond to two different collimation systems and two choices of neutron moderator. The characteristics discussed include the intensity and energy dependence of the neutron flux, the spatial profile of the beam, the in-beam background components and the energy resolution/broadening. The discussion of these features is based on dedicated measurements and Monte Carlo simulations, and includes estimations of the systematic uncertainties of the mentioned quantities. 1 Introduction and objectives The accurate knowledge of a wide variety of nuclear data is essential for the understanding of many of the processes occurring in nature and also playing a key role in nuclear technologies. In particular, neutron-induced reactions are the driving force behind the nucleosynthesis of elements beyond iron in stars (via the s-andr-processes) [1], the functioning of present nuclear reactors [2,3] as well as the future ones aiming at minimizing nuclear waste (Generation-IV) [4] or reducing it (ADS) [5], the damage caused in the structural material of future fusion reactors [6], etc. In all these fields, the cross sections of a large number of neutron-induced reactions are the main ingredient in the associated calculations. The cross sections available in different nuclear libraries (for instance, JEFF, ENDF, JENDL [7] for nuclear technologies and KADoNiS [8] for astrophysics) are derived from evaluations based on both experimental data and theoretical predictions from nuclear models. Nevertheless, the cross sections given in different evaluations are frequently incompatible and their accuracy is lower than needed. This makes it necessary to perform new and more accurate measurements of many isotopes and reactions, which are summarized for instance by the Nuclear Energy Agency in its High Priority Request List [9], but also in review papers on data needs for fusion [10] and astrophysics [1]. The neutron energy ranges of interest span from thermal (25 meV) to hundreds of MeV, depending on the application. In such a wide energy range the neutron cross sections have resonant structures and abrupt reaction thresholds which change significantly from isotope to isotope and cannot be predicted, thus requiring high energy resolution measurements to resolve such structures. This is preferably done by using the Time-of-Flight (ToF) technique in which a pulsed neutron beam, spanning over a wide range of energies, travels a given distance before reaching the sample under study. In this technique the kinetic energy of the neutrons is determined from their time of arrival at the measuring station. The kinetic energy En of neutrons with a speed v=L/t can be expressed as En=E−mc2=c2p2+m2c4−mc2=mc2(γ−1),(1) where γ=(1−v2/c2)−1/2,andcis the speed of light. The first term of the series expansion gives the classical expression for the neutron kinetic energy, which shall be used below a few MeV, En=1 2mv2=α2L2 t2.(2) Taking, for instance, the value of the speed of light c= 299.8m/μsandm= 939.6MeV/c2for the neutron mass, one obtains α=72.29√eVμs m. Regarding the relative energy resolution of a ToF facility, this is given to a first approximation (see sect. 6 for details) by ΔEn En =2×Δt t2 +ΔL L2 ,(3) and therefore the capability for resolving the resonance structures in the cross sections is favored by the use of long flight paths (large Land t) and the production of all neutrons within the shortest possible time (Δt)and smallest possible space (ΔL). At present there are several neutron ToF facilities in the world devoted to the measurement of neutron-induced reactions cross sections. These use different methods for neutron production and have flight path lengths ranging between less than one meter to 400 m. Among these, the most active in the recent years are (flight paths are given in parenthesis) the n TOF facility [11–14] (185 m) discussed further in this paper, GELINA at JRC-IRMM in Belgium [15] (10, 30, 50, 60, 100, 200, 300 and 400 m), DANCE at LANSCE in the USA [16] (20 m) and ANNRI at J-PARK in Japan [17] (22 and 28 m). In addition, two new facilities will soon become operational in Europe: nTOF-EAR2 at CERN in Switzerland [18] (20 m) and FRANZ in Germany [19] (0.8 m). ae-mail: [email protected] Eur. Phys. J. A (2013) 49:27 Page 3 of 15 The n TOF facility became operational at CERN (Geneva, Switzerland) in 2001 and has since then become a major facility in the field of neutron cross section measurements. In addition, measurements aimed at the study of nuclear structure and the mechanism of fission reactions are part of the experimental program. The facility ran between 2001 and 2004 (n TOF-Phase1) and, after a fouryear halt and a complete upgrade of the neutron production target, it resumed operation at the end of 2008 (start of n TOF-Phase2). Since then, many measurements have been performed [20], starting with a series for the commissioning and characterization of the neutron beam, which are discussed in this paper. Everything in this manuscript refers to the n TOF facility after the 2008 upgrade, except when is explicitly specified that something refers to the previous configuration n TOF-Phase1 (2001–2004). The knowledge of every parameter describing the neutron beam is essential for the accurate analysis of experimental data. To this aim, an extensive measurement campaign was undertaken in order to determine these parameters. A significant part of the work presented herein deals with the estimation of the uncertainties of each of the beam parameters and their effect when propagated to the different types of measurements. A brief description of the facility is given in sect. 2, which is followed by a discussion of the Monte Carlo simulations of its performance in sect. 3. The characteristics of the beam discussed in this paper are the flux and its neutron energy dependence (sect. 4), the distribution of the neutron beam in space (sect. 5) and time (sect. 6), and the in-beam background components (sect. 7). 2 Description of the CERN n TOF facility A detailed technical description of the n TOF facility will be given in a dedicated paper (ref. [14]), thus only a brief summary is given in this paper. The n TOF facility is part of the fixed target experimental program at the CERN accelerator complex [21], as sketched in fig. 1. At n TOF, a high-intensity neutron pulse is produced every 1.2 seconds (or multiples of this interval) from spallation reactions induced by a pulse of 7 ×1012 protons of 20 GeV/cmomentum impinging on a 1.3 tonne cylindrical lead target 40 cm in length and 60 cm in diameter, which is illustrated in fig. 2. A 1 cm water layer cools the target and, together with a subsequent layer of 4 cm of water or borated water (H2O+1.28%H3BO3, fraction in mass), moderates the initially fast neutrons into the desired energy spectrum, which ranges down to thermal energies. This separation of the cooling/moderation circuit in two regions is one of the main differences and advantages (see sect. 7.1) with respect to the previous configuration (2001–2004) where a sole water layer of 5.7 cm was used both as coolant and neutron moderator. The experimental area begins at 182.3 m from the spallation target and has a length of 7.9 m. Along the evacuated beam line, sketched in fig. 3, a sweeping magnet (200 cm long, 44 cm gap and 3.6 Tm field) deflects the Fig. 1. Layout of the n TOF facility within the CERN accelerator complex [21]. The LINAC feeds the PS-Booster, which provides the PS with protons of 1.4 GeV/cfor acceleration up to 20 GeV/c. This beam is extracted and sent to the n TOF lead spallation target in bunches of 7×1012 protons. The experimental hall is located near the end of the 200 m long neutron beam line. Fig. 2. Cross section of the n TOF lead spallation target. The separation of the cooling (water) and moderator (water or borated water) layers is realized by a thin alumunium window. charged particles in the beam and two collimators shape the neutron beam. The diameter of the second collimator, placed at 178 m from the spallation target, can be chosen between 18 mm (capture mode with 235 cm of steel plus 50 cm of borated polyethylene) and 80 mm (fission mode with 50 cm of borated polyethylene plus 125 cm of steel plus 75 cm of borated polyethylene) to accommodate the needs of each measurement. Several detection systems are available to study different types of reactions. Liquid (C6D6[22]) and solid (BaF2[23]) scintillators are used to measure the γ-rays following (n,γ) reactions, while gas (MicroMegas [24,25]) and solid-state (silicon [26] and diamond [27]) detectors are employed for (n,charged-particle) reaction measurements. Lastly, fission reactions are measured with gas detectors (MicroMegas [24,25] and PPAC [28]). The combination of these detection systems with a fully digital Data Page 4 of 15 Eur. Phys. J. A (2013) 49:27 Pb Shielding Filter station First collimator Shielding Sweeping magnet Second collimator Experimental area Beam dump Spallation target Fig. 3. Layout of the n TOF neutron beam line from the spallation target to the beam dump (distances are given in meters). Acquisition System [29] has enabled the accurate measurement of a large number of cross sections (the complete list of published results can be found on the n TOF website www.cern.ch/nTOF) and several key technical contributions [22,30–32] to this field of research in the last decade. 3 Monte Carlo simulation of the facility Monte Carlo simulations are used for determining the characteristics of the neutron beam. In particular, once they are validated with experimental data, simulations give access to information that cannot be measured with the required accuracy (experiments always have limitations in terms of statistics and sizeable backgrounds are sometimes unavoidable) or is simply not accessible through experiments (see, for instance, the discussion on the resolution function in sect. 6). This section deals with the technical details of the simulations, while the results and their comparison with experimental data are discussed throughout the corresponding sections later in this paper. 3.1 FLUKA and MCNP simulations Since the start-up of the n TOF facility in 2001, the geometry of the facility has been implemented in FLUKA [33, 34] in considerable detail. However, for neutronics calculations only the region around the spallation target illustrated in fig. 4 is important, since the surrounding structures are only important for estimating the activation of materials, residual dose, etc. The changes made during the upgrade period (2004-2008), particularly the new spallation target assembly, are included in the simulated geometry, which was also exported to MCNP [35] using the builtin feature in FLAIR [36], the FLUKA Graphical User Interface. The choice of using both FLUKA (version 2011) and MCNP (MCNPX version 2.6) was made in order to take advantage of each code’s strong points in a study where a variety of particle interactions, spanning at least 12 orders of magnitude in energy, are followed. In particular, FLUKA boasts very accurate and well-benchmarked high-energy hadron interaction models, appropriate for simulating the proton beam interaction with the spallation target. Nevertheless, the transport of neutrons below Fig. 4. Geometry of the target assembly implemented in the FLUKA and MCNP simulations (partial view). 20 MeV (and down to the lowest threshold of 0.01 meV) is performed by a multi-group algorithm and an associated grouped cross section library containing over 250 materials. The above-mentioned energy range is subdivided into 260 groups of approximately equal logarithmic width where inelastic reactions are not simulated explicitly, but treated as transfer probabilities between energy groups, forming a so-called downscattering matrix. While this approach is generally reliable and very CPU-efficient, it can lead to unphysical artifacts when studying thin and/or low density objects or when the resonance structure of the studied material is relevant to the problem (e.g. in shielding applications). This implies a limitation, for instance, in extracting accurate information on the resonance absorption dips present in the n TOF neutron energy spectrum. It is in this context that MCNP becomes useful, since it employs point-wise neutron cross sections down to thermal energies. 3.2 Neutron production: geometry and scoring The geometry was carefully built based on technical documentation of the n TOF facility (spallation target, vacuum windows, ...) and civil engineering layouts of the environment (tunnels, target area, ...). Furthermore, the Eur. Phys. J. A (2013) 49:27 Page 5 of 15 composition of the materials and especially of the alumunium alloys which constitute the various windows and beam-line components were defined in the best possible detail taking into account the available specifications and their chemical and isotopic composition. The goal of the simulations was to follow the proton beam interaction with the lead spallation target, and to track all particles produced, up to a scoring plane corresponding to the entrance of the neutron beam line vacuum tube. In particular, the neutron or γ-ray position, direction, energy, weight and time of arrival (defined as the time between the primary proton hitting the target and the secondary particles reaching the scoring plane) were recorded for their subsequent transport through the beam line. Particles entering the vacuum tube at an angle greater than 10◦relative to the beam-line axis were immediately discarded as they can never reach the experimental area. All this information was later used to study the neutron propagation towards the experimental area, as described in sect. 3.3. 3.3 Neutron transport along the beam line A full simulation of the neutrons travelling along the beam-line up to the experimental area would be impractical since the solid angle subtended by the second collimator is less than 10−8sr. Such calculations would therefore require an unaffordable long CPU time. To address this issue, an independent external code was developed which uses the information obtained from the simulations to carry out the propagation of the neutrons or γ-rays to the experimental area a posteriori.To this aim, a variance reduction algorithm was implemented. The latter assumes that within a sufficiently small angle θcut (relative to the beam axis) neutrons are emitted isotropically throughout the full energy range. In general, this is not true: high-energy neutron emission is (strongly) forward-peaked, while the angular distribution of thermal neutrons is largely isotropic. Nevertheless, the assumption holds within a small forward angle (θcut =5 ◦), and thus a conservative value of θcut =3 ◦was used for these calculations. The first step is to project each neutron trajectory and determine its hitting point at a distance Lcorresponding to the experimental area (without accounting for any collimation) and discard all those that arrive outside a circle of radius L×tan θcut +0.4m, (e.g. 183 m ×tan 3◦+0.4m =10.0 m), where the 0.4 m radius of the vacuum tube after the lead target (i.e. the neutron scoring plane) is included to account for the beam halo. The remaining neutrons are the ones assumed to be emitted isotropically. Subsequently, a scoring surface in the experimental area is selected, defined by its position along the line (e.g. 183 m) and its radius (e.g. 2 cm). This surface is meant to represent a sample or detector. A scoring grid (1 mm step) is defined on this surface and, for each neutron, a trajectory is calculated to each point on this grid starting from its initial position at the beginning of the vacuum tube. For each instance of the neutron (i.e. for each calculated trajectory) the program checks if the trajectory hits either a tube or collimator (see fig. 3). If so, this neutron is considered not to reach the experimental area and is thus rejected. In the opposite case, it is scored. The final result is appropriately normalized to provide the neutron flux and the spatial beam profile at the scoring surface in the experimental area. In calculating the neutron trajectories, the program was designed to account for the effect of gravity. While fast neutrons are practically unaffected over a ∼185 m flight path, the effect is visible for thermal neutrons (3 cm vertical drop along 185 m) and has a measurable impact on the spatial profile of low energy neutrons (more details are given in sect. 5). 4 Neutron flux The neutron flux is here defined as the number of neutrons per incident proton pulse and integrated over the full spatial beam profile arriving at the experimental hall with a given energy. Its absolute value and energy distribution as well as the associated uncertainties are essential for the analysis of cross section data obtained via the time- of-flight technique. The accuracy with which the neutron flux is determined may indeed become the limiting factor in the accuracy achievable in a cross section experiment and therefore must be improved as much as possible. The knowledge of the dependence of the neutron flux on energy is similarly important as knowing its absolute value because reaction cross sections are usually measured relative to some well known quantity in a particular energy range, e.g. a standard cross section, the cross section value at thermal energies, or the capture yield at the top of a well isolated saturated resonance [37]. In the case of n TOF, we have experimentally determined the neutron flux using several detection systems based on different principles and reactions. These are included in the analysis only in the energy region where the associated cross sections are well-established standards [38]. The measurements, their analysis and their combination for the determination of the evaluated neutron flux are discussed in the following sections. A manuscript with even further details about each measurement and the neutron flux evaluation process is currently under preparation [39]. 4.1 Methodology and measurements The determination of the neutron flux is based on the relation between the measured reaction yield Ymeas xand the expected (theoretical) value Yth xcalculated from the corresponding standard cross section. The former is calculated from measurable quantities as Ymeas x(En)= C(En)−B(En) εx(En)×Φn(En),(4) where Cand Bare the total and background counts per neutron pulse, εxis the detection efficiency for the re- Page 6 of 15 Eur. Phys. J. A (2013) 49:27 action of interest, and Φnis the neutron flux to be determined. All these observables, as explicitly indicated in eq. (4), are neutron energy dependent. The expected yield is calculated, to a first approximation (without considering multiple scattering), from the characteristics of the sample used in the measurement and the associated cross section as Yth x(En)=1−e−nσt(En)σx(En) σt(En),(5) where nis the areal density of the sample in atoms-per- barn and σtand σxare the total and reaction cross sections expressed in barns, respectively. When the sample is sufficiently thin, the self-shielding becomes negligible and thus a combination of the two previous equations allows calculating the neutron flux as Φn(En)= C(En)−B(En) εx(En)×n×σx(En).(6) We have determined the neutron flux of n TOF applying this methodology to several dedicated measurements. By combining these independent measurements in a consistent way, it was possible to determine the neutron flux between thermal and 1 GeV with rather small systematic and statistical uncertainties. The detectors for charged particles SiMon [26] and MGAS [24,25] have been used for measuring (n,α) reactions on thin 6Li and 10B foils, which feature standard cross sections from 25 meV to 1 MeV. In addition, the detectors H19 from PTB [40], MGAS [24,25] and PPAC [28] have all been used to measure 235U(n,f) reactions, whose cross section is standard at 25 meV and in the interval between 150 keV and 200 MeV. This reaction has also been used as reference above 200 MeV, due to the absence of any standard cross section at such high energies. As an illustrative example of the responses of the detectors to the mentioned reactions, fig. 5 shows some pulse height distributions in the neutron energy range between 0.1 and 10 eV. The labels indicate the origin of the structures that are observed in each case and the dashed vertical lines indicate the analysis gates selected for the identification of the reaction products of interest. These gates, shown here only in the eV region, are neutron energy dependent and have been determined carefully for the analysis of the flux in the complete neutron energy range. As mentioned previously, all the measurements performed to determine the neutron flux are based on reactions whose cross sections are standards in particular energy regions [38], as summarized in table 1. The table presents also the energy regions where each detector provides reliable data. For instance, the uncertainties in the angular distribution of the reaction products from the 6Li(n,α) reaction limit the use of the SiMon detector to energies below ∼150 keV, while the effects of the γ-flash (see sect. 7.2) limits the use of the H19(PTB) and MGAS(235U) fission chambers beyond ∼3and∼10 MeV, respectively. In the case of MGAS(10B), the background caused by proton recoils from elastic n-p collisions in the gas limits the high energy range up to only few hundreds of keV. Counts (arb. units) 0 10 20 30 40 1 10 100 Amplitude (flash-ADC units) 100 150 200 250 0 10 20 30 40 050 Counts (arb. units) Counts (arb. units) MGAS(10B) PTB SiMon α 3H α FF 7Li α Saturated signals 235U(n,f)FF 6Li(n,α)3He 10B(n,α)7Li Fig. 5. Distribution of pulse heights corresponding to some of the detectors (and covering the three reaction types of interest) involved in the experimental determination of the n TOF neutron flux. Taking these energy limits into account, the neutron flux has been derived from eq. (6) as described in the following section. 4.2 Evaluated neutron flux To evaluate the neutron flux, the results from each individual measurement have been combined and the overall uncertainty has been estimated consistently. Taking into consideration that the efficiency of the H19 fission detector from PTB and the mass of the corresponding samples are the best known quantities (see ref. [40]), the fluxes measured with the other detectors have been normalized to the PTB detector at 25 meV, where all the measured reactions have standard cross sections. The evaluated neutron flux has then been obtained as a weighted average of the individual fluxes, including in each energy range only those whose associated cross section is standard and considering the reliable energy intervals mentioned in table 1. The results are displayed with a 100 bins per decade in the upper panel of fig. 6, where the wide range of neutron energies provided by the n TOF neutron beam is clearly visible. The two fluxes displayed correspond to the capture collimator mode in the configurations with water and borated water as moderators. In all cases the MeV Eur. Phys. J. A (2013) 49:27 Page 7 of 15 Table 1. Summary of the detectors used for the determination of the neutron flux, including the reactions of interest, the energy range in which their cross section is considered standard, the sample thicknesses, the collimator used, and the (approximate) upper energy limit for proper operation of each detector. Detector Sample Areal density Collimation Enwhere σxUpper Enlimit and reaction (μg/cm2) (diameter in mm) is standard for detection(a) PTB 235U(n,f) 500 (10 samples) Capture (18) 25 meV, 0.15–200 MeV 3 MeV SiMon 6Li(n,t) 300 Capture (18) 25 meV to 1 MeV 150 keV MGAS 10B(n,α) 55 Capture (18) 25 meV to 1 MeV 150 keV MGAS 235U(n,f) 470 Capture (18) 25 meV, 0.15–200 MeV 1 MeV PPAC 235U(n,f) 279 Fission (80) 25 meV, 0.15–200 MeV 1 GeV (a) See text for details. Neutron Flux (EdΦ/dE/Pulse) 3 10 4 10 5 10 Neutron energy (eV) -2 10 -1 10 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 Uncertainty (%) 0 1 2 3 4 5 Capture collimator (18 mm) H20 H20+10B Statistical Systematic Dips due to aluminum along the beam line Fig. 6. Top: Evaluated neutron flux in the configuration with the capture collimator (18 mm diameter) with water and borated water as moderator. Bottom: Systematic and statistical uncertainties (see text for details). region shows the typical neutron evaporation spectrum, with the intermediate region containing the partially moderated neutrons and the low-energy region, corresponding to thermalized neutrons, changing significantly when borated water is used instead of normal water due the higher neutron absorption probability through 10B(n,α) than through 1H(n,γ) reactions. This also leads to a reduction of the intensity and energy of the in-beam γ-ray background, as discussed in sect. 7.1. The region between 20 keV and 1 MeV features a series of dips, which are transmission dips associated to resonances in the alumunium windows of the spallation target assembly and at both ends of the evacuated flight path. These resonance dips have a sizable effect when measuring the cross section of isotopes with resonances at such high energies, but Table 2. Number of neutrons per pulse, assuming a nominal pulse intensity of 7 ×1012 protons per pulse), in different neutron energy ranges in the capture and fission collimator modes using water and borated water as moderator. Neutron Neutrons/pulse energy Capture mode Fission mode H2OH 2O+10BH 2O+10B 10–100 meV 1.0×1051.5×1043.3×105 0.1–1 eV 4.3×1041.3×1042.9×105 1–10 eV 2.7×1042.0×1044.3×105 10–100 eV 2.8×1042.5×1045.3×105 0.1–1 keV 2.9×1042.9×1046.2×105 1–10 keV 3.2×1043.2×1046.8×105 10–100 keV 4.4×1044.4×1049.4×105 0.1–1 MeV 1.3×1051.3×1052.7×106 1–10 MeV 1.5×1051.5×1053.1×106 10–100 MeV 5.0×1045.0×1041.0×106 0.1–1 GeV 4.7×1044.7×1049.5×105 Total 6.8×1055.5×1051.2×107 only slightly affect the measurement of actinides and their unresolved resonance cross section in this region. The number of neutrons per pulse (corresponding to 7×1012 protons) integrated over energy decades are given for each collimator and moderator configurations in table 2. The total number of neutrons per pulse varies between 5.5×105and 12 ×106, depending on the collimator/moderator configuration. Regarding the statistical and systematic uncertainties in the neutron flux, displayed in the botton panel of fig. 6, the former has been calculated through standard uncertainty propagation and the latter has been estimated as the Root Mean Square (RMS) of the values that are included in the averaging at each neutron energy bin. In this way, the lower limit for the systematic uncertainty is given by the statistical one. Indeed, it is observed that the systematic uncertainty is larger than the statistical only Page 8 of 15 Eur. Phys. J. A (2013) 49:27 near the large alumunium dip around 20–30 keV. Overall, the neutron flux with 100 bins per decade has been determined within 1% uncertainty between thermal and 100 eV, 2% between 100 eV and 10 keV, 3–4% between 10 keV and 1 MeV, and 2–3% between 1 and 200 MeV. Above 200 MeV the 235U(n,f) cross section is not standard and therefore we can only report the statistical uncertainty of our flux, which varies between 3 and 4%. These uncertainty values constrain the maximum accuracy that can be reached in absolute cross section measurements at nTOF; however, the results from both fission and capture cross section measurements can reach higher accuracies if performed relative to standard cross sections such as 197Au(n,γ)or235U(n,f). 4.3 Consequences of a high instantaneous flux Many of the measurements carried out at n TOF are made on radioactive samples such as actinides (233U, 240,242Pu, 241,243Am, 245Cm,...),fissionproducts (93Zr, 151Sm,...) and others (59Ni, 63Ni, ...). In these cases, the activity of the sample itself often becomes the dominant source of background. At n TOF, due to the very high instantaneous intensity of each neutron pulse, the contribution of the sample-activity background to the measurement is largely reduced with respect to other facilities. As an example fig. 7 shows the counting rate recorded with a pair C6D6scintillators in the (n,γ) measurement of a32mg241Am sample [41,42]. The activity of this sample was 4 GBq, and even with such a high activity the measurement could be carried out at n TOF with a favorable capture-to-background ratio in the complete energy range. The example also serves to illustrate the wide neutron energy range, between thermal and 1 MeV, fully covered in each pulse at n TOF thanks to its uniquely short duty cycle. Neutron energy (eV) 1 10 1 10 2 10 3 10 4 10 5 10 6 10 Counts (arb. units) 4 10 5 10 6 10 7 10 241Am (Beam On) 241Am (Beam Off) 241Am activity Fig. 7. Neutron energy distribution corresponding to the measurement of (n,γ) reactions from a 32 mg 241Am sample with an activity of 4 GBq. Beam size (cm) -4 -2 4 Normalized distribution 0 0.2 0.4 0.6 0.8 1Meas. (18 mm collim.) MC (18 mm collim.) MC (80 mm collim.) 02 Fig. 8. Neutron beam profile at 183.2 m in the vertical direction as measured with the pixel-MGAS detector at low neutron energy (0.1–1 eV). 5 Spatial beam profile The spatial profile of the neutron beam at the sample position is mainly shaped by the characteristics of the second collimator. The two collimation modes available at n TOF (see sect. 2) obviously provide neutron beams of different size. Furthermore, the beam profile varies with the energy of the neutrons because neutrons of different energies are produced at different positions and emitted at different angles from the spallation target and thus follow different paths during their moderation in the target and the cooling/moderator system. The knowledge of this spatial profile is essential for the accurate measurement of cross sections and is of particular importance when the sample being measured is smaller than the neutron beam and therefore intercepts only a fraction of it, referred to as the Beam Interception Factor (BIF). The spatial profile of the neutron beam and its energy dependence in the capture collimator configuration were already investigated in detail in n TOF-Phase1 by means of a 1D-stripped MicroMegas detector (see ref. [43] for details). The results were found to be in agreement with simulations available at the time. In the present n TOF configuration, a new 2D pixelated MicroMegas (pixel-MGAS) detector has been developed [44]. This new detector, with a 5 cm diameter active area including 308 square pixels of 2.5 mm side read out by 4 Gassiplex cards, has provided in 2012 results compatible with those measured in 2001 with the stripped-MGAS detector. An illustration of the n TOF beam profile at 183.2 m flight path is given in fig. 8, where the beam projection in the vertical direction for neutron energies between 0.1 and 1 eV is shown for both capture and fission collimator modes. The capture mode includes both the measurement with the pixel-MGAS and simulations (see sect. 3.1), which are in good agreement except in the tails of the dis- Eur. Phys. J. A (2013) 49:27 Page 9 of 15 tribution, where the measured data are affected by background. In the fission mode, only the simulations are available. It is observed that the full size of the beam is ∼3.5 cm and ∼10 cm in diameter in the capture and fission collimation modes, respectively. 5.1 Beam interception factor The value of the BIF at 183.2 m flight path using the capture collimator has been calculated from the validated simulations for typical sample diameters of 1, 2, 3 and 4 cm, always considering samples that are perfectly aligned with the beam line. The results are shown in fig. 9, where the BIF values in the eV region are 0.21, 0.68, 0.96 and 1.0 for samples with diameters of 1, 2, 3 and 4 cm, respectively. Figure 9 clearly shows the slight variation of the BIF with neutron energy which needs to be accounted for when the cross section measurements are normalized in a particular neutron energy region. The difference between the BIF value at 4.9 eV (which is the usual normalization point for capture experiments at n TOF via the saturated resonance method [37] with 197Au) and the values at thermal and at 300 keV reaches 2.5% for samples of only 1 cm diameter, and is less than 1.5% for samples with a diameter larger than 2 cm. The figure also includes 4 lines around the BIF values for 2 cm diameter that correspond to the values for a sample misaligned by 1 mm upwards, downwards, left and right. These results show that an accurate sample aligment is crucial, since a misaligment of just 1 mm can lead to a deviation of several percentage points with respect to the BIF value at the nominal sample position. Neutron energy (eV) -2 10 -1 10 1 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 Beam Interception Factor 0 0.2 0.4 0.6 0.8 1 1 cm diameter 2 cm diameter 3 cm diameter 4 cm diameter Fig. 9. Simulated beam interception factor BIF (at 183.2 m flight path) as a function of neutron energy for circular samples 1 to 4 cm in diameter. The four thin lines around the 2 cm diameter values correspond to sample misalignments of ±1mm horizontally and vertically. The effect is indeed larger in the case of a vertical misalignment, which is explained by the effect of gravitation on thermal neutrons, since it takes them more than 100 ms to reach the experimental area. Overall, it can be concluded that the variation of the BIF with neutron energy must be considered in data analysis when measuring samples smaller than the beam, considering as well that any possible misalignments of the sample could imply an uncertainty on the neutron energy dependence of the BIF, and thus in the corresponding measured cross section, which can reach a maximum of 2.5% at thermal energies for samples with a diameter of only 1 cm. 6 Resolution function and ToF/Enrelation The resolution function describes the distribution of the measured time of flight for neutrons with a given kinetic energy. Several experimental conditions like the primary beam pulse duration (7 ns RMS in the case of n TOF), the moderation time in the target-moderator assembly, and the detector response, contribute to the fact that the time of flight tand the flight length Lfor a neutron of energy Enin eq. (2) are not fixed values but merely distributions. The convolution of these distributions is known as the resolution function RE(En) for a given incident neutron energy and can be easily transformed in equivalent distributions in either energy, time, or distance by using the conservation of probability RE(En)dEn=Rt(t)dt=RL(L)dL, (7) where En,Land tare related to each other through eq. (2). Since the energy dependence of the equivalent distance RL(L) is weaker than Rt(t)andRE(E), this form is often used to represent resolution functions graphically. Although the resolution function cannot be measured directly, it is assessed either by an analytical approach of its components or, more commonly, by Monte Carlo simulations. The resolution function can then be validated by measurements of well-known resonances. The resolution function is usually non-Gaussian and even asymmetric which does not only alter the shape of resolved resonances, but also changes the time-energy calibration with energy, as a consequence of a shift in the observed resonance peak position. For resolved resonances analyzed with R-matrix codes like SAMMY [45] or REFIT [46], a parametrization of the resolution function or a numerical representation is used to fit the resonance shape and its position. The combination of an arbitrary geometrical flight length Land the also arbitrary center of the resolution function results in the time-energy calibration. Usually, but not necessarily, the resolution function starts at zero equivalent distance, and the geometrical flight length is calibrated accordingly with a well-known low-energy resonance. For reaction yields at higher energies where no resolved resonances are present, the conversion of time to energy with eq. (2) uses the effective flight length Leff which