Precise Neutron Lifetime Measurement Using Pulsed Neutron Beams at J-PARC
N. Sumi, K. Hirota, G. Ichikawa, T. Ino, Y. Iwashita, S. Kajiwara, Y. Kato, M. Kitaguchi, K. Mishima, K. Morikawa, T. Mogi, H. Oide, H. Okabe, H. Otono, T. Shima, H. M. Shimizu, Y. Sugisawa, T. Tanabe, S. Yamashita, K. Yano, T. Yoshioka
PPrecise Neutron Lifetime Measurement Using PulsedNeutron Beams at J-PARC
N. SUMI , K. HIROTA , G. ICHIKAWA , T. INO , Y. IWASHITA , S. KAJIWARA , Y.KATO , M. KITAGUCHI , K. MISHIMA , K. MORIKAWA , T. MOGI , H. OIDE , H.OKABE , H. OTONO , T. SHIMA , H. M. SHIMIZU , Y. SUGISAWA , T. TANABE ,S. YAMASHITA , K. YANO and T. YOSHIOKA Department of Physics, Kyushu University, Fukuoka 819-0395, Japan Research Center for Nuclear Physics, Osaka University, Ibaraki, Osaka 567-0047, Japan KEK, High Energy Accelerator Research Organization, Tsukuba 305-0801, Japan Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan Department of Physics, The University of Tokyo, Bunkyo-Ku, Tokyo 113-0033, Japan Kobayashi-Maskawa Institute for the Origin of Particles and the Universe, Nagoya University,Nagoya 464-8602, Japan Department of Physics, Nagoya University, Nagoya 464-8602, Japan Department of Physics, Tokyo Institute of Technology, Tokyo 152-8551, Japan Research Center for Advanced Particle Physics, Kyushu University, Fukuoka 819-0395, Japan Institute of Applied Physics, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan International Center for the Elementary Particle Physics, The University of Tokyo, Tokyo113-0033, JapanE-mail: [email protected] (Received January 10, 2020)A neutron decays into a proton, an electron, and an anti-neutrino through the beta-decay process. Thedecay lifetime ( ∼
880 s) is an important parameter in the weak interaction. For example, the neutronlifetime is a parameter used to determine the | V ud | parameter of the CKM quark mixing matrix. Thelifetime is also one of the input parameters for the Big Bang Nucleosynthesis, which predicts lightelement synthesis in the early universe. However, experimental measurements of the neutron lifetimetoday are significantly di ff erent (8.4 s or 4.0 σ ) depending on the methods. One is a bottle methodmeasuring surviving neutron in the neutron storage bottle. The other is a beam method measuringneutron beam flux and neutron decay rate in the detector. There is a discussion that the discrepancycomes from unconsidered systematic error or undetectable decay mode, such as dark decay. A newtype of beam experiment is performed at the BL05 MLF J-PARC. This experiment measured neutronflux and decay rate simultaneously with a time projection chamber using a pulsed neutron beam. Wewill present the world situation of neutron lifetime and the latest results at J-PARC. KEYWORDS: neutron, lifetime, MLF, J-PARC
1. Introduction
A free neutron decays into a proton, electron, and anti-neutrino with a mean lifetime τ n ∼
15 mindenoted as, n → p + e − + ν e . (1)Figure 1 shows the measured neutron lifetime in these twenty years. There are two types of methods,one is called “storage method” and the other is “beam method”. The discrepancy between these twomethods of 8.6 s or 4.1 σ is called “neutron lifetime anomaly”. Before explaining the measurementmethods in detail, the physical significance of τ n will be introduced in the next section. a r X i v : . [ h e p - e x ] F e b ig. 1. The measured neutron lifetime over the publication year. There are two types of methods, one iscalled “storage method” and the other is “beam method”. The discrepancy between these two methods of 8.6 sor 4.1 σ is called “neutron lifetime anomaly”. The Big Bang Nucleosynthesis (BBN) is a theory that estimates the production of the light ele-ment in the early universe. Since the time scale of the BBN is similar to τ n , the abundance of lightnuclei strongly depends on it. Figure 2 is the observations of the early universe and the prediction ofhelium abundance Y p = He / (H + He). The predicted Y p is the cross point of the band of τ n andbaryon to photon ratio η , which is determined by the Planck satellite from the observation of cosmicmicrowave background (CMB) [1]. There are two bands of τ n by the measurement methods. Twoobservations (Aver:2015 [2] and Valerdi:2019 [3]) are in good agreement with the prediction, but oneobservation (Izotov:2014 [4]) does not. Since the observed accuracy of Y p and η is improving year byyear, the ambiguity of τ n should be resolved. Baryon to photon ratio0.2 0.3 0.4 0.5 0.6 0.7 0.8 -9 · P r i m o r d i a l H e li u m a bund a n ce Y p – = 879.4 n t Storage: 2.0 sec – = 888.0 n t Beam:
Planck:2018
Valerdi:2019Aver:2015Izotov:2014
Fig. 2.
The observations and the prediction of helium abundance Y p . The three filled rectangular regions areobserved results of Y p . The vertical region is the baryon to photon ratio η and the two curved bands are theprediction of BBN by the two τ n results. 2 .2 Unitarity of CKM matrix In the standard model of particle physics, the Cabibbo-Kobayashi-Maskawa (CKM) matrix de-scribes the strength of transitions between quarks in weak interactions. The unitarity check of theCKM matrix gives a strong test of the standard model. For example, the first row of the matrix gives | V ud | + | V us | + | V ub | = . ± . | V ud | comes fromthe study of superallowed J π = + → + nuclear beta decays, which are pure vector transitions. Theerror of | V ud | is dominated by theoretical uncertainties stemming from nuclear Coulomb distortionsand radiative corrections. A precise determination of | V ud | is also obtained from the measurement ofneutron decay as | V ud | = (4908 . ± .
9) s τ n (1 + λ ) . (2)The theoretical uncertainties are very small, but the determination is limited by the uncertainties of theratio of the axial-vector and vector couplings, λ = g A / g V , and τ n . Figure 3 shows | V ud | values alongwith λ . The filled black box indicates | V ud | that satisfies unitarity. The hatched and filled magentaboxes indicate | V ud | by superallowed nuclear decay with an old and a new radiative correction [6],respectively. This value has a slightly smaller value from the unitarity with the correction. The valueobtained from the neutron decay is the cross point of τ n and λ . The cross point of τ n by the storagemethod and λ by Perkeo III [7] and that of the beam method and a SEPCT [8] have | V ud | value closeto the unitarity. The value λ can be calculated from the QCD lattice gauge theory, but the calculatedresults cannot reproduce the experimental value [9]. Fig. 3. | V ud | values along with λ . The filled black box indicates | V ud | that satisfies unitarity. The hatched andfilled magenta boxes indicate | V ud | by superallowed nuclear decay with an old and a new radiative correction [6],respectively. The value obtained from the neutron decay is the cross point of τ n and λ . The cross point of τ n bystorage method and λ by Perkeo III [7] and that of beam method and a SEPCT [8] have | V ud | value close to theunitarity. Note that, results by the neutron lifetime are also a ff ected by the radiative correction. To explain the disagreement between “storage method” and “beam method”, Fornal et al. suggestthat the neutron cloud decay into unobserved particles by 1% of the usual β decay [10]. In the beammethod, experimentalists could not observe unexpected decay mode as neutron decay and the result ot longer. On the other hand, in the storage method, the result would not rely on the decay mode.They propose that the neutron could decay into dark matter particles with the following decay modes, n → χ + γ (937 .
900 MeV < m χ < .
783 MeV) (3) n → χ + e + + e − (937 .
900 MeV < m χ < .
543 MeV) (4) n → χ + φ (937 .
900 MeV < m χ + m φ < .
565 MeV) , (5)where φ is another dark matter particle. The mass of dark matter m χ and m φ are strictly limited bythe stability of the proton and nuclei. After the publication of the neutron dark decay paper, someexperiments [11, 12] rejected some decay modes.
2. Measurement methods
The storage method measures neutron lifetime by storing ultracold neutron (UCN) in the specificbottle. They counts the number of surviving neutrons S and S after distinct storing times t and t .Then, τ n is calculated by, ln ( S / S ) t − t = τ n + τ wall . (6)In this equation, τ wall is the wall loss e ff ect of the stored neutron. There are many reasons to loseneutrons from the bottle, e.g. absorption and scattering. The estimation and correction of the τ wall is the key point of the storage method. In this big gravitational trap experiment [13], the ultracoldneutrons were guided and filled into the UCN trap. After a certain storing time, the survived neutronswere released to the neutron detector below the bottle. The τ wall was estimated 1.5% for the lifetimeof the neutron by changing the volume or temperature of the bottle. This experiment published theresult of τ n = . ± . ± . τ n = . ± . + . − . (syst) s and τ n = . ± . ± . The beam method measures neutron lifetime by counting the injected neutron and decay productin the beam. In this penning trap experiment [16], the neutron beam was injected into the volume andthe decay proton was stored in the magnetic field and electric field. The flux of the injected neutronbeam was monitored by converted to charged particles at a thin Li plate via, Li + n → α + t. Then,these α -ray or triton was detected by the surrounding detectors. After the neutron beam was stopped,the trapped protons were accelerated toward the proton detector along the magnetic field by one sideof the electrode voltage was dropped to 0 V. The neutron lifetime was obtained from the counting ratioof these two detectors. This experiment published the result of τ n = . ± . ± . . In another beam method using Time Projection Chamber (TPC) [17], neutron lifetime is obtainedfrom the simultaneous measurement of an electron from β decay and He(n,p) H reaction. Theychopped neutron beam to short bunch and injected them into TPC. The neutron lifetime was obtainedfrom, τ n = ρσ v (cid:18) S He /ε He S β /ε β (cid:19) (7)where S β and S He are the counting numbers of β decay and He(n,p) H reaction, ε β , ε He are the de-tection e ffi ciency of them, v is velocity of the neutron, ρ and σ are the number density and absorptioncross section of He. This experiment published the result of τ n = ±
27 (stat) ±
14 (syst) s . Itsaccuracy was limited by the statistics and the background for the β decay signals. . Neutron lifetime measurement at J-PARC Figure 4 is a schematic view of the neutron lifetime measurement at BL05 [18] in the Mate-rials and Life Science Experimental Facility (MLF), Japan Proton Accelerator Research Complex(J-PARC). The neutron beam is chopped at the spin flip chopper (SFC) to make a short bunch and in-jected it to the TPC. The neutron shutter is installed at the upstream of the TPC to control the neutronbunch. The TPC counts the β decay and He absorption signals. The rest of the bunch is absorbed inthe beam dump.
Spin FlipChopper TPC 1m(D)(A)(a) (b) (c) (d) (e)(1) (2) (3) (4) (5) (6) (7) (8)(B) (C)(Z)(Y)(X) zx BL05 BL04
Fence C on c r e t e S h i e l d o f B L04
BL06 (constructing)
Fig. 4.
Schematic view of the neutron lifetime measurement at BL05. (X) Polarized (Y) High intensity (Z)Low divergence beam branch (A) Concrete shield (B) Lead shield (C) Iron shield (D) LiF neutron duct (a)Collimator (b) Guide coil (c) Spin Flipper (d) Magnetic super mirror (e) Neutron beam monitor (1) Zr window(2) LiF neutron shutter (3) Cosmic veto counter (4) Lead shield (5) Vacuum chamber (6) Time ProjectionChamber (7) Beam dump (8) Vacuum pump
The neutron beam data have been acquired since 2014. In this paper, the data from 2014 to 2016are used to analyze. These acquired data are summarized in table I.
Gas Date MLF power [kW] Beam time [hour]I May 2014 300 35.3II April 2015 500 15.8III April 2016 200 17.5IV April 2016 200 72.7V May 2016 200 69.4VI June 2016 200 71.1
Table I.
Acquired data from 2014 to 2016. One data set is corresponding to one gas fill. MLF power denotesJ-PARC proton beam power to which the neutron beam power is proportional.
The number of two signals S β , S He are extracted from the acquired data using signal selectioncut. The time-independent backgrounds are subtracted using the time of fight method and the neutronshutter open and closed data. The cut e ffi ciencies ε β , ε He , and the number of remaining backgroundsare estimated using Monte Carlo simulation. Then, the neutron lifetime τ n can be calculated by equa-tion 7. .1 Signal selection The selection for the neutron β decay signal is the following five cuts. The first one is “time offlight cut” which requires that event trigger time is in the neutron bunch completely inside the TPC( −
380 mm < z <
380 mm). The second is “drift length cut” which requires that drift length, or y length, is smaller than half of the TPC ( <
190 mm). The third is “distance from beam axis cut” whichrequires that the edge of the track on the beam axis within ±
48 mm. The fourth is “point like cut”which requires that the range of the track is greater than 100 mm or deposit energy is greater than5 keV to eliminate CO recoil point-like event. The last is “high energy cut” which requires that theenergy on a low gain wire is smaller than 25 keV for all wires to eliminate He absorption.The selection for the neutron He absorption signal is following two cuts. The first one is “timeof flight cut” which is the same as β decay signal. The second one is an inversion of “high energycut” which requires that any of the low gain wire exceeds 25 keV. The time-independent background is subtracted using time of flight method and neutron shutteropen and closed data. The β decay and neutron-induced background emerge only while the neutronbunch in the TPC, defined as “fiducial time”. Ahead of it, upstream γ -ray coming from SFC generatesbackground peak and more background comes from the mercury target at the time of flight T = F ( τ / = Li ( τ / =
840 ms). Since they have a longer lifetime than the MLF beam cycle of 40 ms, they are regarded astime constant background. However, their lifetimes are shorter than an interval of the shutter openand closed, they disappear at closed operation.
Time of Flight [ms]0 5 10 15 20 25 30 35 40 c oun t / . m s / s tof Shutter openShutter closedSubtructed Fig. 5.
Time of flight drawn by acquired data Gas II. The red and black filled areas are the neutron shutteropen and closed state and the blue area is a subtraction of them. There are five clear peaks on the small flatbackground in the subtraction spectrum.
The remaining backgrounds for S β are estimated using Monte Carlo simulation. In the CO capture reaction C + n → C (1keV) + γ (4945keV), the recoiled C deposits small energyon the beam axis, but 99.9% events are eliminated by “point like cut”. However, if the γ -ray scatterselectron in TPC, it turns to the background. The scattered neutron β decay is treated as background ecause they have an unpredictable track for each decay point. The wall capture γ -ray is the dominantsource of the remaining background. The detector wall captures scattered neutron and emits γ -ray. Ifthe γ -ray scatter an electron, it turns to the background.Figure 6 shows estimation of the background contamination over distance from beam axis. Thedistance = >
120 mm represents trackedge on the wall. The black point indicates experimental data. The stacking histograms are simulationdata, scattered β decay (orange), wall capture γ -ray (blue), and β decay (red) from the bottom. Thesimulation data were scaled to the o ff -axis control region. The estimated contamination is 463 ± capture reaction is negligible. Distance from beam axis [mm]0 12 24 36 48 60 72 84 96 108 120 132 C oun t Fig. 6.
Background contamination for S β over distance from beam axis. The distance = >
120 mm represents track edge on the wall. The black point indicateexperimental data. The stacking histograms are simulation data, scattered β decay (orange), wall capture γ -ray(blue), and β decay (red) from the bottom. The backgrounds for S He are following two absorption on the beam axis, N + n → C + p + , (8) O + n → C + α + . (9)Although they have a di ff erent energy from He absorption of 764 keV, the gain of Multi Wire Pro-portional Chamber (MWPC) is saturated at their energy region in high gain operation. Therefore,low gain operation data were taken once a day to measure contamination of them. The source ofnitrogen is outgassing in the vacuum chamber and its rate was estimated at 0.4 Pa / day and correctedits contribution. The source of oxygen is CO as the quencher gas of 15 kPa. Unlike N, fluctua-tion by outgassing is negligible, thus its contribution was calculated by the cross section and naturalabundance as a time constant and corrected 0.50% for He of 100 mPa. Besides, He absorption ofthe scattered neutron is treated as background same as scattered neutron β decay. The contribution ofsuch an event was evaluated by 0.30% by the amount of o ff -axis He absorption.
4. Result and uncertainty
Table II is results and uncertainties of the all values in equation 7 for the Gas II data. The numberof β decay signal S β has the largest uncertainty in the values due to background contamination. Thee ffi ciency of β decay signal ε β has the next largest uncertainty. The number density of He gas willbe improved to 0.1% with the updated injection method. Another experiment is planned to improvethe accuracy of the cross section σ of He. ρ (2.08 ± / m He gas injection σ v ± × / s 0 0.13% Requires other measurement S He ±
480 2672 ±
351 0.3% Statistical error S β ±
151 463 ±
154 2.6% Background contamination ε He (100 − ε β (94.5 ± + ± Table II.
The results and uncertainties for the Gas II.
The combined result of all Gas I - VI is, τ n = ±
10 (stat) + − (syst) s . (10)This result has still large uncertainty to compare the other results, but it is consistent with the beammethod and storage method. Since this method is independent of the other methods, the improvedresult gives a hint to discuss the disagreement.In addition, we published an updated result, τ n = ±
10 (stat) + − (syst) s , (11)by reconsidering the systematic uncertainty [20]. Acknowledgements
This research was supported by JSPS KAKENHI Grant Number 19GS0210 and JP16H02194.The neutron experiment at the Materials and Life Science Experimental Facility of the J-PARC wasperformed under a user program (Proposal No. 2015A0316, 2014B0271, 2014A0244, 2012B0219,and 2012A0075) and S-type project of KEK (Proposal No. 2014S03).