Detection capability of Migdal effect for argon and xenon nuclei with position sensitive gaseous detectors
Kiseki D. Nakamura, Kentaro Miuchi, Shingo Kazama, Yutaro Shoji, Masahiro Ibe, Wakutaka Nakano
aa r X i v : . [ phy s i c s . i n s - d e t ] S e p Prog. Theor. Exp. Phys. , 00000 (13 pages)DOI: 10.1093 / ptep/0000000000 Detection capability of Migdal effect for argonand xenon nuclei with position sensitivegaseous detectors
Kiseki D. Nakamura , Kentaro Miuchi , Shingo Kazama , Yutaro Shoji ,Masahiro Ibe , and Wakutaka Nakano Department of Physics, Graduate School of Science, Kobe University, 1-1Rokkodai-cho, Nada-ku, Kobe, Hyogo, 657-8501, Japan ∗ E-mail: [email protected] Racah Institute of Physics, Hebrew University of Jerusalem, Jerusalem 91904, Israel Institute for Cosmic Ray Research (ICRR), The University of Tokyo, Chiba277-8583, Japan Kavli Institute for the Physics and Mathematics of the Universe (WPI), TheUniversity of Tokyo Institutes for Advanced Study, The University of Tokyo,Kashiwa 277-8583, Japan Department of Physics, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan
August 2020 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Migdal effect is attracting interests because of the potential to enhance the sensitivitiesof direct dark matter searches to the low mass region. In spite of its great importance,the Migdal effect has not been experimentally observed yet. A realistic experimentalapproach towards the first observation of the Migdal effect in the neutron scatteringwas studied with Monte Carlo simulations. In this study, potential background rate wasstudied together with the event rate of the Migdal effect by a neutron source. It wasfound that a table-top sized ∼ (30cm) position-sensitive gaseous detector filled withargon or xenon target gas can detect characteristic signatures of the Migdal effect withsufficient rates (O(10 ∼ ) events/day). A simulation result of a simple experimentalset-up showed two significant background sources, namely the intrinsic neutrons and theneutron induced gamma-rays. These background rates were found to be much higherthan those of the Migdal effect in the neutron scattering. As a consequence of this study,it is concluded that the experimental observation of the Migdal effect in the neutronscattering can be realized with a good understanding and reduction of the background.
1. Introduction
Revealing the nature of the unidentified gravitational source in the universe, or the darkmatter, is one of the most important task in particle physics, astroparticle physics andcosmology. Weakly Interacting Massive Particles (WIMPs) are said to be one of the mostattractive candidates and have been searched for decades by various ways [1, 2]. In spiteof these intensive searches for the “standard” WIMPs with a mass of hundred GeV toseveral TeV, they have not been discovered yet. Although the interest and motivation forthese standard WIMPs have not decreased [3], some experimental efforts to search for other c (cid:13) The Author(s) 2012. Published by Oxford University Press on behalf of the Physical Society of Japan.This is an Open Access article distributed under the terms of the Creative Commons Attribution License(http://creativecommons.org/licenses/by-nc/3.0), which permits unrestricted use,distribution, and reproduction in any medium, provided the original work is properly cited. ypes of dark matter candidates are being carried out. Among many types of other darkmatter candidates [4, 5], lighter WIMPs are attracting interests and some experiments haveinvestigated this new frontier [6–8].Low mass WIMPs are conventionally searched for with low threshold WIMPs detectorssuch as bolometers and “S2-only analysis” of two-phase liquid noble gas detectors. Recently,some groups reported the results and plans for light WIMPs searches using the effect socalled the Migdal effect [9–14] motivated by the recent reformulation of the effect [15] (seealso [16, 17]).The Migdal effect refers to immediate excitation or ionization of the recoil atom at thestandard nuclear recoil, which was first calculated in a different context by A. B. Migdal in1941 [18]. A sudden change of the nuclear velocity caused by the recoil disturbs the electricpotential of the atom. When its energy is absorbed by the surrounding electrons, the recoilatom is either excited or ionized. Then, the excited/ionized atom immediately de-excitesthrough the X-ray transition, the Auger transition, or the Coster-Kronig transition, whichdeposits additional energy inside the detector. The rate of having the Migdal effect dependson the momentum transfer, but it is very suppressed in the case of light WIMPs.These Migdal effects can even take place in a very low energy nuclear recoil where therecoil atoms are hardly detectable due to the poor ionization capability. Thus the Migdaleffect effectively lowers the detection energy threshold of nuclear recoils and drasticallyimproves the sensitivity of the low mass WIMPs searches. In spite of its great importancefor the WIMPs searches, the Migdal effect itself has not been observed experimentally yet.A realistic experimental approach, including the potential background sources, towards thefirst observation of the Migdal effect is discussed in this paper.
2. Conceptual experimental design
As introduced in the previous section, the experimental observation of the Migdal effect itselfis important for the low mass WIMPs searches. We hereby propose a method to detect thecharacteristic signature of the Migdal effect; a nuclear recoil with an additional electric signal.In order to detect the topology of these two signals, position-sensitive gaseous detectors wereconsidered. Among three possible reactions of the electric signals, we focused on the reactionassociated with the emission of the characteristic X-rays so that the electric signal can bespatially-separated from the nuclear recoils with an appropriate choice of the gas and itspressure. This “two-cluster” event topology can be used to select signals out of huge amountof background events to claim a clear observation of the Migdal effect. Argon gas at 1 atmand xenon gas at 8 atm are chosen for this study so that the characteristic X-ray absorptionlength would be several cm.Time projection chambers (TPCs) are widely-used position-sensitive gaseous detectors.Micro Pattern Gaseous Detectors (MPGD) are well-studied readout devices for gaseousargon TPCs around the normal pressure. A typical three-dimensional spatial resolution ofmm or less and an energy resolution of about 30% FWHM at 5.9 keV are realized [19, 20]. Areadout system using the electroluminescence (EL) photons are developed for a high pressuregaseous xenon detector [21]. The energy resolution of 4% at 30 keV has been demonstratedwith the EL signal read by photon detectors with a pitch of 10 mm [21]. Considering theserealistic detectors, we assume a detector with detection volume of (30 cm) with MPGD and L readouts for the argon and xenon TPCs, respectively. Total numbers of the target nucleifor the argon and the xenon detectors are shown in Table 1.There are various neutron sources potentially available for this experiment. In this studya continuous neutron beam with an energy of 565 keV and a flux of 1000 / cm / sec at 1 mis assumed. This is a typical value for a Li(p , n) Be reaction at an irradiation facility inNational Institute of Advanced Industrial Science and Technology (AIST), Japan. The crosssections of elastic scattering between 565 keV neutrons and the target nuclei are shown inTable 1. It should be noted that, in general, the higher the neutron energy is, the moregamma-ray background rate from the detector and the laboratory materials is expected.Thus we choose this energy as a typical neutron energy instead of more common sources likeDT (deutron-triton) or DD (deutron-deutron) generators.
3. Signal
In this section, Migdal effect signals are discussed. First, an overview of the features of theMigdal effect including the expected event rate is given based on Ref. [15]. Then a morerealistic MC simulation study on the detection possibility of the characteristic features ofthe Migdal effect is described.
Probabilities of the Migdal effects depends on the quantum numbers ( n, l ) of the associatedelectron. When an electron with a large principal quantum number ( n ) is involved, theprobability of the Migdal effect is large but the additional electric energy is small because ofthe small binding energy. Observations of the Migdal effects with large principal quantumnumbers, especially with n ≥ n, l ) = (1 ,
0) cases (1s, K shell) for both argon and xenon whereadditional electric energies can be detected with practical detectors are chosen for the firstobservation of the Migdal effect. The branching ratios of the Migdal effect for the ( n, l ) =(1 ,
0) cases are listed in Table 1.Figure 1 illustrates the reactions we expect to observe. The first reaction is the nuclearrecoil (1) with a recoil energy of E NR and the Migdal effect (2). The orbital electron thatreceives energy by the Migdal effect transitions to another orbit or goes out of the atom.Since the probability of latter process is several orders of magnitude higher than the formerone, the latter process is studies in this paper. The recoil nuclei and the Migdal electronare detected as one cluster (cluster A). A de-excitation takes place following the emission ofthe Migdal electron. The de-excitation can be seen as a characteristic X-ray (3) or an Augerelectron. Probabilities of de-excitation of K-shell hole by X-ray emission, or the fluorescenceyield, are listed in Table 1. Characteristic X-ray can be detected at a distance of severalcm with the detector condition we assume so that a second cluster (cluster B) with a fixedenergy ( E dex ) is expected. This topology information will provide us with a clear signal ofthe Migdal effect together with a powerful background rejection. Finally, multiple Augerelectrons and de-excitation X-rays with a total energy of E nl − E dex are emitted for theenergy conservation, where E nl is the binding energy of the ( n, l ) electron (4). ig. 1 Shcematis mechanism of the reactions related to the Migdal effect.At the leading order, the probabilities of the Migdal effect are known to be proportionalto the square of the momentum transfer to the Migdal electron q e . Probabilities for a given q e are calculated by scaling the values shown in Table 2 of Ref. [15] with ( q e /
511 eV) . q e iscalculated as follows, q = 2 m E NR m N , (1)where m e is the electron mass and m N is the target nuclear mass. Table 1 shows the scalingfactors for maximum recoil energies E maxNR by an irradiation with 565 keV neutrons. Here E maxNR is known by E maxNR = 4 m n m N ( m n + m N ) E n . (2)Here m n is the neutron mass and E n is the neutron energy. It is seen that the scaling factorfor argon is one order of magnitude larger than that for xenon because m N is smaller and E NR is consequently larger for a given energy of neutrons.Expected event rates calculated for the experimental and physical conditions discussedso far are shown in the final row of Table 1. Here the event rate for the nuclear recoilsassociated with characteristic X-rays from the Migdal effect are shown. The rates ( O (10 ∼ ) events/day) themselves without the consideration of any backgrounds are encouragingones. In reality, the background rates without any reduction are much larger than thesesignal rates. It is one of the important points of this study to discuss a realistic methodto discriminate background events with the event topologies, which will be discussed inSection 4.It should be emphasized that our calculation and discussed measurement are only for theisolated atoms. For the application to the dark matter searches, it is also important to testthe Migdal effect in the liquid medium. A simple calculation showed an encouraging signal event rate as an ideal case. A morerealistic Geant4 [22] Monte Carlo (MC) simulation study was then performed.Four particles, listed as (1) ∼ (4) in the followings, are generated for the MC simulation. arget gas Ar 1 atm (30 cm) Xe 8 atm (30 cm) number of nuclei 7 . × × cross section for 565 keV neutron 0.65 barn 6.0 barnMigdal branching 7 . × − . × − fluorescence yield (K shell) 0.14 0.89scaling factor ( q maxe /
511 eV) Table 1
Typical values of parameters for estimating the Migdal effect. The branchingratios for ( n, l ) = 1s and q e = 511 eV are shown.(1) A recoil nucleus with a recoil energy of E NR following the probability distributiondiscussed in Ref. [15] is generated. The position is homogeneously distributed in thedetection volume. Here E NR , recoil angle θ and the transition energy to the electron(∆ E ) have correlations as shown in equation (4.18) in Ref. [15].(2) A Migdal electron with an energy E e = ∆ E − E nl is generated.(3) A de-excitation X-ray with an energy E dex is generated. K α or K β X-ray with knownintensity ratio is generated with its characteristic energy.(4) De-excitation electron with an energy E nl − E dex is generated.The initial directions of the particle generated in (2)-(4) are isotropic.Figure 2 and 3 show the results of the signal simulations for argon and xenon targets,respectively. The total energy spectra, energy spectra of the X-rays (cluster B), and thedistributions of the distance between two clusters are shown in the left, center, and rightpanels, respectively. Here the events with any energy deposition within 1 cm from the wallwere rejected (fiducial cut) in order to reject the background charged particles from the wallof the detector. Another criteria for the event selection was to require the distance betweentwo clusters to be larger than 1 cm.The left plots show the quenching factor dependence of the total energy. Three cases withthe quenching factors of 1, 0.5, and 0.1 are shown. The case of the quenching=1, where therecoil nuclei would show the same ionization efficiency with electrons, is shown as a reference.Since the quenching factor depends on the detector parameters such as the density of themedium, impurities, and the electric field, it needs to be measured at an actual detectorcondition. The real quenching factors can be between 0.1 and 0.5, and thus energy spectrafor these two cases are shown as references. The real spectrum would be expected somewherebetween these two spectra. The effect of quenching appears more clearly in the spectrumwith argon target. This is because the nuclear recoil energy is larger than the energy of thecharacteristic X-ray. Hereafter we will discuss with the case of quenching=1, i.e. withoutconsidering the quenching. This would help to understand the real recoil energy and can becertified since the event selections can be tuned with a measured quenching factor prior tothe Migdal effect experiments.In the center plots, it is seen that the energy spectra of cluster B make peaks, so that thebackground can be efficiently rejected by requiring two clusters with either one having theenergy of the characteristic X-ray. It should be noted that the quenching factors need notto be considered for the X-ray signals. he distance distributions between two clusters shown in the right plots have exponentialshape representing the absorption length of X-rays. The best-fit curves shown with red linesare consistent with the ones for the expected absorption length (2.95 cm and 2.19 cm forargon and xenon, respectively). This shape would be a very good evidence of the Migdaleffect in contrast to the flat distribution for the neutron multiple scatterings or square tothe distance for accidental backgrounds. c oun t s / da y / k e V total energy quenching=1quenching=0.5quenching=0.1 Migdal of Ar1s by 565keV neutronAr 1atm c oun t s / da y / k e V energy of cluster-B − c oun t s / da y / c m distance between clusters fitted by f(x)=A*exp(-x/X)X=2.72cm Fig. 2
Signal MC simulation results with argon gas. The spectra of the total energy,energy spectra of cluster B, and the distributions of the distance between the two clustersare shown in the left, center, and right, respectively. c oun t s / da y / k e V total energy quenching=1quenching=0.5quenching=0.1 Migdal of Xe1s by 565keV neutronXe 8atm c oun t s / da y / k e V energy of cluster-B − c oun t s / da y / c m distance between clusters fitted by f(x)=A*exp(-x/X)X=2.29cm Fig. 3
Signal MC simulation results with xenon gas. The spectra of the total energy,energy spectra of cluster B, and the distributions of the distance between two clusters areshown in the left, center, and right, respectively.
4. Background
Background studies were carried out with Geant4 MC simulations. In this study, it turnedout that there exist two significant background sources, namely the intrinsic neutrons andthe neutron induced gamma-rays. An unavoidable intrinsic neutron background caused bythe incident neutron on the target gas will be discussed first. Then, a gamma-ray backgroundfrom the (n , γ ) reaction in the chamber and the laboratory material will be evaluated. .1. Intrinsic neutron background
Nuclear recoils by the incident neutrons associated with another energy deposition willbecome unavoidable background events because their event topology would look like theMigdal signal events discussed in the previous section.First, accidental coincidence background where two independent nuclear recoils take placeswithin a time resolution of the TPC ( ∼ µ s) is studied. The probabilities of accidentalcoincidence are O (10 − ) and O (10 − ) for the argon and xenon targets, respectively. Since thecluster distance of accidental coincidence event follow quadratic function, a further O (10 − )reduction is expected. For the argon case, the event rate is smaller than that of the Migdalsignal. For the xenon case, the accidental coincidence can be rejected by requiring the energyof cluster B to be 30 keV since the maximum recoil energy is about 18 keV.We then studied the background events created by interactions of neutrons and targetnuclei, mainly multiple scatterings and inelastic scatterings, as an intrinsic background. Thekey parameters to reject these intrinsic backgrounds are the energy of the cluster B andthe distance between two clusters as shown in the center and right panels of Figure 2 andFigure 3. The rate and the rejection efficiency of the background events were studied withthe MC simulations. In order to study the pure effect of this intrinsic background, a detectorwith a gas-only target of (30 cm) was prepared and irradiated with neutrons of the sameenergy (565 keV) as the signal study.Figure 4 shows the total energy spectra obtained by the argon-target MC simulations. Thespectrum without any cut is shown with the black-filled histgram (cut 0). The spectrum afterthe fiducial cut, same as the signal selection, is shown with the blue-filled histogram (cut1). Elastic scatterings make dominant contribution below 60 keV. The background rate isfound to decrease by about 4 orders of magnitude by requiring two clusters (cut 2, red-filled histogram). The background rate above 60 keV is further decreased by requiring thepresence of a cluster with an energy of characteristic X-ray (3.2 ± × cm and the shape of the distribution is determined by the detectorsize. The background distance distribution (blue) is in good agreement with the efficiencycurve of detecting two clusters in a (28 cm) volume. The cluster-distance distribution ofthe Migdal effect events is found to be more steep, showing a shorter absorption length ofthe characteristic X-ray. Therefore, the observation of the Migdal effect can be claimed byextracting the signal rate from the cluster-distance distribution.Figure 6 shows the total energy spectra obtained by the xenon-target MC simulations.Cut criteria 0 to 2 are same as the ones for the argon-target case. Cut 3 requires that one
10 20 30 40 50 60 70 80 90 100keV −
10 110 c oun t s / da y / k e V intrinsic neutron BG cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 5.2keV cluster cut3: 1.2keV < EMigdal (cut3) Ar 1atm
Fig. 4
Simulated total energy spectra ofthe intrinsic neutron background events forthe argon target. Black-filled histogram isthe raw energy spectrum. Blue, red, andgray ones are those after cut 1, 2, and 3,respectively. The black-solid line is the energyspectrum of the Migdal effect. c oun t s / da y / c m distance between clusters expected SIG+BGSIG: Migdal (cut3)BG: intrinsic neutron (cut3) Ar 1atm
Fig. 5
MC simulation results on the dis-tance between two clusters for the argon tar-get. Red, blue, and black histogram show theintrinsic neutron background events, signalevents and sum of these two components.of the clusters has the energy of the K α X-rays of xenon, 29.6 keV ± Xe (natural abundance 26.4 %) which has an inelastic scattering associated with40 keV or 120 keV γ -rays. These inelastic scatterings make two clusters, one by the nuclearrecoil and another by the X-ray, and make the discrimination difficult even with the eventtopologies. Next-largest contribution is Xe (natural abundance 21.2 %) which has anemission of 80 keV γ -rays. Although the contribution of Xe is much smaller than that of
Xe, it is still larger than the count rate of the Migdal effect. The third one comes from
Xe (natural abundance 4.1 %) which is associated with high energy ( ∼ MeV) gamma-rays.Partial energy depositions of Compton-scattered electrons from these gamma-rays make thecontinuum component of the background spectrum. Other isotopes make background lessthan the Migdal effect. Among the various isotopes, it is seen that
Xe and
Xe whichhave natural abundances of 10.4 % and 8.9 % makes less background than the Migdal effect.In this MC study, the background contribution of
Xe which is produced by the neutroncapture of
Xe and emits 81 keV prompt gamma-rays is not taken account because of along half-life of 5.2 days. Since the rate is estimated to be around O (10 ) counts/day, it isnecessary to apply energy selection of <
80 keV or remove
Xe by isotope separation. he possibility of the Migdal effect observation with xenon gas becomes realistic with anisotope enrichment of the isotopes larger than A=133. More than four orders of reductionfor
Xe and
Xe are required considering the original contributions of these isotopes tothe background spectrum. With an additional energy selection between 60 keV and 80 keV,the Migdal effect can be observed with xenon gas. −
10 110 c oun t s / da y / k e V intrinsic neutron BG cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 31.1keV cluster cut3: 28.1keV < EMigdal (cut3) Xe 8atm
Fig. 6
Simulated total energy spectra ofthe intrinsic neutron background events forthe xenon target. Black-filled histogram isthe raw energy spectrum. Blue, red, andgray ones are those after cut 1, 2, and 3,respectively. The black-solid line is the energyspectrum of the Migdal effect. −
10 110 c oun t s / da y / k e V isotope ratio of neutron BG neutron BG (cut3)Xe Xe Xe Xe Xe, Xe Xe, Xe Xe,
Fig. 7
Isotope-breakdown of the intrin-sic neutron background events for the xenontarget.
Neutrons interact with materials of the laboratory and the detector and generate gamma-raysvia (n, γ ) reactions. The production rates and the energies of gamma-ray backgrounds dependon the laboratory geometry and detector design. Here, typical gamma-ray backgrounds for asimplified geometry shown in Fig. 8 are studied. A stainless chamber with a thickness of 5 mmis implemented enclosing a gas medium with a volume of (36 cm) . The detection volume iskept same as the one used in the signal simulation (30 cm) while a 3 cm buffer is preparedfor each plane. The neutron source is set at a position 1 m away from the detector center,same condition as the signal simulation. For the gamma-rays from the laboratory materials,neutrons from the (p,Li) reaction are generated to directions taking account of the energyand rate dependence. The energy-angle correlation is show in Fig. 9. The detector is setat the center of a laboratory with a size of (11 . . An aluminum floor with an effectivethickness of 7.8 mm is constructed 1 m below the detector center.The MC results are shown in Figs. 10 ∼
13. The selection criteria are same as the onesused in the intrinsic neutron backgrounds. The gamma-ray backgrounds from the chambermaterial for the argon and xenon cases are shown in Fig. 10 and Fig. 11, respectively. The ig. 8
A geometry for the gamma-raybackground MC simulation. The figure is notto scale.
LAB θ ene r g y [ M e V ] angle dependence of neutron energy Fig. 9
Angular dependence of neutronenergy due to Li(p , n) Be reaction. Blackdots are simulation results by AIST, red linesare calculation results from kinematics. Theneutrons used to estimate the gamma raybackground due to the (n , γ ) reaction withthe structure were generated according to thered line.chamber has the gas medium inside so that these background spectra are the sum of theintrinsic neutron background and the gamma-ray background. The gamma-ray backgroundsfrom the chamber material are shown with green-filled histograms for a better understanding.It is seen that the gamma-ray background rate is smaller than that of the intrinsic neutronbackground for the argon case. The gamma-ray background rate is also found to be smallerthan that of the intrinsic neutron background but significantly larger than that of the Migdalsignal for the xenon case. Possibilities to reduce the chamber background are discussed inSection 4.3.The gamma-ray background spectra from the laboratory material for the argon and xenoncases are shown in Fig. 12 and Fig. 13, respectively. (n, γ ) reactions take places in theconcrete and aluminum floor and the gamma-ray flux at the detector position is recorded.The gamma-rays are newly generated at the detector position to increase the statistics. Therate of the gamma-ray backgrounds from the laboratory are found to be much larger thanthat of the Migdal effect. These backgrounds need to be reduced at the generation point ofthe neutrons. Hydrocarbon and B shiledings with a hole only to the detector direction canbe set around the reaction point to reduce the number of neutrons reaching the laboratorymaterials. The gamma-rays from the laboratory can also be reduced with shieldings aroundthe detector. It should be noted that these shieldings need to be designed with a goodconsideration of the laboratory geometry and neutron energy because these new materialwould be new sources of the gamma-ray background. With these shieldings, the gamma-raybackground rates need to be reduced at least two orders of magnitude to observe the Migdaleffect.
10 20 30 40 50 60 70 80 90 100keV −
10 110 c oun t s / da y / k e V ) BG + neutron BG γ chamber (n, cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 5.2keV cluster1 cut3: 1.2keV < Ecut4: no NRMigdal (cut3) Ar 1atm
Fig. 10
Results of the gamma-ray back-ground MC simulation study. The energyspectra from the chamber material for theargon gas detector are shown. The colors aresame as the ones used in Fig. 4 with anadditional green-filled histogram showing thegamma-ray background. −
10 110 c oun t s / da y / k e V ) BG + neutron BG γ chamber (n, cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 31.1keV cluster cut3: 28.1keV < Ecut4: no NRMigdal (cut3) Xe 8atm
Fig. 11
Results of the gamma-ray back-ground MC simulation study. The energyspectra from the chamber material for thexenon gas detector are shown. The colors aresame as the ones used in Fig. 10. −
10 110 c oun t s / da y / k e V ) BG γ laboratory (n, cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 5.2keV cluster cut3: 1.2keV < EMigdal (cut3) Ar 1atm
Fig. 12
Results of the gamma-ray back-ground MC simulation study. The energyspectra from the laboratory material for theargon gas detector are shown. The colors aresame as the ones used in Fig. 4 −
10 110 c oun t s / da y / k e V ) BG γ laboratory (n, cut0: no cutcut1: fiducial cut = 2 cluster cut2: N < 31.1keV cluster cut3: 28.1keV < EMigdal (cut3) Xe 8atm
Fig. 13
Results of the gamma-ray back-ground MC simulation study. The energyspectra from the laboratory material for thexenon gas detector are shown. The colors aresame as the ones used in Fig. 6. .3. Discussion
The background study so far assumes a continuous (DC) neutron beam. It would be possibleto reduce delayed or continuous background such as laboratory gamma-rays with a pulsedneutron beam. The effect of the background rejection with the pulsed neutron beam woulddepend on the time evolution of the background rate. One needs to have a good understand-ing of the neutron beam to estimate the background rejection by a pulsed-beam. Gammaray background from the chamber can be reduced by designing a chamber with a smalleramount of material like polyimide film for the argon gas case because the gas pressure is atthe normal pressure. For the xenon gas case, on the other hand, chamber material cannot bereduced because of its high pressure. Self-shielding technique can be used instead by increas-ing the detector volume and use outer ∼
10 cm as a veto region. Also, it may be effective toinstall a solid scintillator on the wall inside the chamber [23] as an active shieldings. In thissense this study is giving a conservative estimation of the background rate.The angular distribution of the Migdal electron from n = 1 follows cos θ N − e , with θ N − e as the polar angle from the momentum of the recoil nucleus. Therefore, if the directionof the nucleus and the direction of the Migdal electron can be measured, the backgroundcan be reduced further. This effect would be valid even by measuring only the directionof the Migdal electron because the direction of the recoil nuclei is biased in the forwarddirection. The challenging point of this method is that the energy of the Migdal electronsis about 10 keV resulting a track length of 2 mm and 1 mm in the argon and xenon gas,respectively. In the future, it would be interesting to be able to use the direction informationof Migdal electrons by some breakthrough of the technology like emulsions, low-pressuregaseous detectors, or columnar recombination.
5. Conclusions
A possibility of experimental observation of the Migdal effect in the neutron scattering withposition sensitive gaseous detectors is studied. Some modes of the Migdal effects would havetwo clusters, one is for the nuclear recoil and the other is for the associated X-ray. Table-top sized position-sensitive gaseous detectors ∼ (30cm) filled with argon or xenon targetgas are found to be capable of detecting sufficient rates (O(10 ∼ ) events/day) of thesetwo-cluster Migdal events. Distributions of the distance between two clusters, representingthe absorption length of the characteristic X-rays, would be a strong evidence of the Migdaleffect observation.Two significant background sources, namely the intrinsic neutrons and the neutron inducedgamma-rays are found to exist. These background rates are found to be higher than theMigdal effect rates even with a relatively low-energy (565 keV) neutron source. It is foundthat the background rate can be reduced to the same order of the Migdal rate with areasonable shielding against the gamma-rays from the laboratory for the argon case. Theintrinsic background by inelastic scatterings would be a serious problem for the xenon case,which can be overcome with isotope separations. Remaining gamma-ray background fromthe chamber and the laboratory can be reduced with precisely-designed shieldings. As aconsequence of this study, it is concluded that the experimental observation of the Migdaleffect can be realized with a good understanding and reduction of the background. cknowledgment This work was supported by KAKENHI Grant-in-Aids (18K13567, 26104005, 16H02189,19H05806, 18H03697, 18H05542, 19H05810), ICRR Joint-Usage, I-CORE Program of theIsrael Planning Budgeting Committee (grant No. 1937/12), World Premier InternationalResearch Center Initiative (WPI), MEXT, Japan.
References [1] F. Donato, Physics of the Dark Universe, , 41 – 43 (2014).[2] M. Schumann, Journal of Physics G: Nuclear and Particle Physics, (10), 103003 (2019).[3] M. Saito et al ., The European Physical Journal C, (6), 469 (2019).[4] T. Braine et al., Phys. Rev. Lett., , 101303 (2020).[5] M. Fabbrichesi, E. Gabrielli, G. Lanfranchi, arXiv:2005.01515 (2020).[6] R. Agnese et al., Phys. Rev. Lett., , 061802 (2018).[7] A. H. Abdelhameed et al., Phys. Rev. D, , 102002 (2019).[8] P. Agnes et al., Phys. Rev. Lett., , 081307 (2018).[9] E. Aprile et al., Phys. Rev. Lett., , 241803 (Dec 2019).[10] D. S. Akerib et al., Phys. Rev. Lett., , 131301 (2019).[11] E. Armengaud et al., Phys. Rev. D, , 082003 (2019).[12] Z. Z. Liu et al., Phys. Rev. Lett., , 161301 (2019).[13] L. Barak et al ., arXiv:2004.11378v1 (2020).[14] G. G. di Cortona, A. Messina, S. Piacentini, arXiv:2006.02453 (2020).[15] M. Ibe et al ., Journal of High Energy Physics, (3), 194 (Mar 2018).[16] M. J. Dolan, F. Kahlhoefer, C. McCabe, Phys. Rev. Lett., , 101801 (2018).[17] R. Essig et al ., Phys. Rev. Lett., (2), 021801 (2020), arXiv:1908.10881.[18] A. B. Migdal, J. Phys. USSR, , 449 – 454 (1941).[19] T. Nagayoshi et al ., Nuclear Instruments and Methods in Physics Research A, A 513 (1-2), 277 – 281(11 2003).[20] K. Miuchi et al ., Progress of Theoretical and Experimental Physics, (6), 063H01 (06 2019).[21] S. Ban et al ., Progress of Theoretical and Experimental Physics, (3), 033H01 (03 2020).[22] J. Allison et al., Nuclear Instruments and Methods in Physics Research Section A: Accelerators,Spectrometers, Detectors and Associated Equipment, , 186 – 225 (2016).[23] T. Tanimori et al ., The Astrophysical Journal, (1), 28 (aug 2015).(1), 28 (aug 2015).