β -delayed neutron spectroscopy of 85 , 86 As with MONSTER

. The β -delayed neutron emission in the 85 , 86 As β -decays has been measured at the IGISOL facility of the Accelerator Laboratory of the University of Jyväskylä (JYFL-ACCLAB). The complete β -decays have been studied with a complex setup that consists of a plastic scintillator, the MO dular N eutron time-of-ﬂight S pectrome TER (MONSTER), and two types of γ -ray detectors—an HPGe clover and four LaBr 3 crystals. The β -delayed neutron energy distributions have been determined by unfolding the TOF spectra with an innovative methodology based on the iterative Bayesian unfolding method and accurate Monte Carlo simulations.


Introduction
The field of β-delayed neutrons has experienced an increased activity during the last decades [1] thanks to the advances in nuclear experimental techniques and radioactive ion beam (RIB) facilities producing increased yields of neutron-rich isotopes.Properties from individual precursors like the emission probability, β-feeding, and energy spectrum are being measured with advanced neutron detectors [2][3][4], digital data acquisition (DAQ) systems [5], and high intensity RIBs [6][7][8][9].
In this paper, the results obtained in the measurement at the IGISOL facility of the Accelerator Laboratory of the University of Jyväskylä (JYFL-ACCLAB) on the βdelayed neutron energy distributions following the 85,86 As β-decays are reported.Some relevant properties of the βdecays of interest are presented in Table 1 [10].

Experimental setup
The 85,86 As isotopes were produced by proton-induced fission reactions in a nat U target.In each case, the isobars were separated from the bulk of fission products by the IGISOL dipole magnet with resolution M/ΔM = 500 [9] and implanted on a movable tape inside of the β-detector.
The β-detector [11] consisted of a BC408 cup-like shaped plastic detector wrapped with reflector foil and coupled to an R5924-70 PMT from Hamamatsu.The detector was placed inside the vacuum pipe and had a 10 mm aperture on the side of the detector cup that allowed for the implantation on the tape to take place.
The emitted β-delayed neutrons were detected with the MOdular Neutron time-of-flight SpectromeTER (MON-STER) [4,12].MONSTER consists of cylindrical cells of 200 mm diameter and 50 mm height, filled with either BC501A or EJ301 scintillating liquid.Each cell was coupled through a light guide of 31 mm thickness to either a R4144 or R11833 PMT from Hamamatsu.For this experiment, 48 cells arranged in two different setups, with 30 and 18 cells at 2 and 1.5 m flight paths, respectively, were used.
The signals from the β-detector and MONSTER were used as the start and stop signals, respectively, to measure the neutrons' time-of-flight (TOF).A picture of the whole ensemble can be seen in Figure 1.All the detector signals were registered with a custom digital DAQ system developed at CIEMAT based on ADQ14DC digitizers from Teledyne SP Devices with 14 bits vertical resolution, 1 GSample/s sampling rate, and 4 channels [5].The complete setup required 60 channels.The system integrates custom pulse shape analysis software also developed at CIEMAT to analyze the registered signals online.The used pulse shape analysis routines did not add any dead time to the measurement.
The data was collected in cycles of implantation and decay adapted to the half-lives of the isotopes involved for maximizing the activity of the neutron emitters with respect to the other implanted isobars.During the cycle, the ion beam was switched on and off by electrostatic deflection at the switchyard of the IGISOL beam line.The cycle also included an initial time interval for background measurement and a final time interval for removal of the long-lived activity of the decay products.

Efficiency characterization
The β-detector efficiency was determined experimentally at two different endpoint energy values from the 92 Sr into 92 Y β-decay through the β-γ coincidence method using the HPGe and LaBr 3 detectors.The Monte Carlo simulations were used for extending the detector efficiency as a function of the endpoint energy and for estimating the detection threshold of 260 keV.The results of the efficiency calculation and the comparison with the experimental data are shown in Figure 2.
The MONSTER neutron detection efficiency was determined through Monte Carlo simulations and validated with the data from a measurement with a 1.00 ± 0.15 GBq 252 Cf source.The γ-rays from the spontaneous fission of 252 Cf detected in the LaBr 3 detectors served for establishing a TOF between the emission and the detection of the spontaneous fission neutrons.Neutron signals in MON-STER were selected by performing pulse shape discrimination, allowing for a reduction of the background due to γ-rays of one order of magnitude.The response of MON-STER to different neutron energies was obtained by setting different cuts in the γ-neutron TOF spectrum.The Monte Carlo simulations were performed with the light output response obtained from previous calibrations with monochromatic energy neutron beams [12].There is an

β-activity distributions
The time evolution of the β-activity during each cycle due to all isotopes in the decay chain is governed by the Bateman equations.The experimental β-activity distribution is fitted with a function based on the generic solution given in [13] that describes the contribution of each of the isotopes involved in the decay chain: where i is the average efficiency to detect a β-particle coming from the β-decay of the i-th member of the chain, λ i is the decay constant of the i-th member of the chain, and N i (t) is the number of nuclei of the i-th member of the chain given by the solution of the Bateman equations, which depends on the implantation rate R i .
The experimental β-activity distributions for both decays, together with the total fits and the individual contributions, can be seen in Figure 4.The measurement cycle in the case of 86 As was shorter in order to maximize its contribution with respect to the other implanted isobars.The results of the fits are presented in Table 2.The other values that are in the table are the calculated average detection efficiencies as well as the total number of decays of each of the isotopes.The average detection efficiencies for each of the β-decays were calculated through Monte Carlo simulations using available data.

TOF spectra unfolding
An innovative analysis procedure based on the iterative Bayesian unfolding method [14] was developed to obtain the neutron energy distributions from the measured TOF spectra [15].
Assuming an initial probability distribution of independent causes (neutron energies) and the probability of each of the causes to produce specific effects (TOF spectrum), the conditional probability of a specific effect (E j ) being due to a certain cause (C i ) is given by: Table 2. Results of the fit to the β-activity curve, along with the calculated average efficiencies and the total number of decays.

Isotope
The conditional probability P(E j |C i ) is the response matrix of the detection system and can be estimated through Monte Carlo simulations.The construction of the response matrix is crucial to achieve an accurate result.The built response matrices covered the whole energy range up to the Q βn values of the decays in intervals chosen according to the energy resolution of the system, and included an extra flat contribution to account for the γ-ray background.The measured experimental TOF spectra for both de-Figure 5 for the MONSTER at 2 m.
Finally, the neutron energy distributions obtained with both MONSTER arrays are presented in Figure 6.The agreement of the results obtained with both arrays is excellent within their energy resolution.The obtained results are also compared to existing data [10,16].The existing data has been rebinned to the binning of the results of this work.Also, the obtained energy distributions have been scaled so that the intensities over the experimental threshold of around 300 keV are equal to those of the existing data.The results obtained for 85 As are in excellent agreement with evaluations, which include experimental data obtained by the Mainz group and are supplemented with theoretical calculations.It's worth noting that no neutron intensity over around 3 MeV, which is predicted by calculations, is observed.No previous experimental data has been found for 86 As on the bibliography.The results of this work are in reasonable agreement with evaluations.

Summary
This work is a successful commissioning and measurement with MONSTER, a neutron TOF spectrometer with excellent neutron/γ-ray discrimination capabilities and neutron energy resolution.
The efficiency of the different detectors has been obtained experimentally and reproduced with accurate Monte Carlo simulations.
An innovative analysis methodology has been developed.It is based on the fit of the β-activity curves with the Bateman equations and on the unfolding of the neutron TOF spectra with the iterative Bayesian unfolding method based on accurate Monte Carlo simulations.
The β-delayed neutron energy spectrum of 85 As has been procured.This result is in excellent agreement with previous experimental data and evaluations, and serves as the validation of the developed analysis methodology.
Lastly, the β-delayed neutron energy spectrum of 86 As has been procured for the first time.The obtained energy distribution is in reasonable agreement with evaluations.

Figure 4 .
Figure 4. Time evolution of the β-activity distributions and the corresponding fits, including individual contributions.

Figure 5 .
Figure 5. Experimental TOF spectra together with the reconstructions and individual causes' contributions.

Figure 6 .
Figure 6.Neutron energy distributions obtained with both MONSTER arrays and compared to existing data.

Table 1 .
Relevant properties of the β-decays.