Project of Neutron Beta-Decay A-Asymmetry Measurement with Relative Accuracy of (1–2)×10–3

Volume 110, Number 4, July-August 2005 Journal of Research of the National Institute of Standards and Technology

[J. Res. Natl. Inst. Stand. Technol. 110, 383-387 (2005)] Project of Neutron Beta-Decay A-Asymmetry Measurement With Relative Accuracy of (1–2)×10–3

Volume 110 Number 4 July-August 2005

A. Serebrov, Yu. Rudnev, We are going to use a polarized cold neu- the background. The final accuracy of the A. Murashkin, O. Zherebtsov, tron beam and an axial magnetic field in A-asymmetry will be determined by the the shape of a bottle formed by a super- uncertainty of the neutron beam polariza- A. Kharitonov, V. Korolev, conducting magnetic system. Such a con- tion measurement which is at the level of T. Morozov, A. Fomin, figuration of magnetic fields allows us to (1–2) × 10–3, as shown in previous studies. V. Pusenkov, A. Schebetov, and extract the decay electrons inside a well- V. Varlamov defined solid angle with high accuracy. An electrostatic cylinder with a potential of 25 Key words: beta-decay; cold neutrons; kV defines the detected region of neutron polarization. PNPI, St. Petersburg Nuclear decays. The protons, which come from this Physics Institute, region will be accelerated and registered Accepted: August 11, 2004 188300, Gatchina, Russia by a proton detector. The use of coincidences between electron and proton sig- nals will allow us to considerably suppress Available online: http://www.nist.gov/jres

1. Introduction The coefficient A is the most sensitive to the λ-value, which is the ratio of fundamental coupling constants An improved measurement of the neutron β-decay ∆∆λ A (λ = G /G ). A =–2λ(λ + 1)/(1 + 3λ2); = 0.25 A-asymmetry is extremely important to test the A V 0 λ A . Standard Model of interaction of elementary particles. Using the precise measurement of the neutron lifetime

In the framework of the Standard Model of V-A-weak and A-asymmetry it is possible to determine the Vud ele- interactions the probability of neutron β-decay is writ- ment of the CKM matrix. The analysis of unitarity of ten in the following form: CKM from the work [1] and a new result for the neutron lifetime [2] require improved accuracy for the A-  =+v asymmetry measurement. The most precise experiment WE(,,)eeppvv fE ()1 e a cos(,) pp e  λ c [3] gives A0 = –0.1189(7) and = –1.2739(19). The  required level of accuracy has to be 2 × 10–3 or better to ++v σ σ APcos( ,ppvv ) BP cos( , ) , (1) c  reach the comparable accuracy of Vud determination from neutron β-decay and from high quark generation ∼ β where W(Ee, pe, pv) is the probability of neutron - decay. decay; Ee, pe are the energy and momentum of the elec- We are going to reach the necessary level of accura- ∼ tron; pv is the momentum of the antineutrino; v is the cy using a new experimental scheme for the A-asym- velocity of the electron; a, A, B are the correlation coef- metry measurement. ficients; σ is the neutron spin; P is the neutron beam polarization.

383 Volume 110, Number 4, July-August 2005 Journal of Research of the National Institute of Standards and Technology

2. The Scheme of the Proposed equal to zero in this case. Therefore we can measure the Experiment A correlation coefficient without any correction for admixtures of other correlation coefficients. The sig- The scheme of the correlation spectrometer is shown nals from the electron and proton detectors have to be in Fig. 1. The correlation spectrometer is itself a sole- detected in coincidence to fulfill this condition. noid with a uniform magnetic field and a magnetic plug Additionally the use of a coincidence regime of regis- at one of its ends. Such a magnetic system acts as a tration of β-decay events will allow us to considerably magnetic collimator for decay electrons. Electrons suppress the background, and to separate effect and whose momentum with respect to the axis are inside the background by means of the delayed coincidence θ solid angle C can go through the magnetic plug field method. and reach the electron detector. Electrons outside this To detect the protons which have very low initial solid angle will be reflected by the magnetic field of the energy, the decay region is placed at high voltage (25 θ plug. The angle C is defined by the ratio of the mag- kV). When protons leave the high voltage region they netic field in the uniform part and in the magnetic plug, are accelerated up to the energy of 25 keV and they can and it is not dependent on the energy of electron be registered by the proton detector. The electrostatic system of the correlation spectrometer consists of a 2 θ = sinC HHOm / , (2) multi-section cylinder with very thin nets at the ends. The sections of the electrostatic system are under dif- where HO is the magnetic field value in the region of ferent potentials from 21 kV up to 26 kV. Due to the neutron decay (uniform part) and Hm is the magnetic gradient of high voltage potential an electric field field value in the magnetic plug. appears inside the cylinder that is aligned along the axis The long solenoid L = 3320 mm produces a magnet- of the spectrometer. All protons will be accelerated in ic field HO = 0.35 T. The short solenoid produces a the direction of the proton detector independently of magnetic field Hm = 0.87 T. For this configuration of their initial momentum. This scheme of the electrostat- θ magnetic fields C is equal to 39 °. The electron and ic system allows us to collect all the protons and to proton detectors are installed on the axis of the magnet- obtain zero average cosines of a and B correlations. The ic system. The region of detected β-decay events is protons can be collected during a time of less than 10 defined by the intersection of the neutron beam with the µs. This helps to reduce the background of random magnetic field lines that go through the electron detec- coincidences. tor. This region is crosshatched in Fig. 1. The size of the The main task of the experimental procedure is the proton detector has to be large enough to collect all pro- measurement of the experimental asymmetry which is tons related to the detected electron. Fig. 2 shows the defined in the following way: magnetic field lines and trajectories of electrons and NN↑− ↓ v protons around the lines. The condition of 100 % col- XAP==cos(σ ,p ) (3) ↑+ ↓ ne lection of protons is very important as the average NN c cosines for the a and B correlation coefficients will be

Fig. 1. Scheme of correlation spectrometer.

384 Volume 110, Number 4, July-August 2005 Journal of Research of the National Institute of Standards and Technology

Fig. 2. Magnetic field and trajectories of electrons and protons. (Scales for y and z directions are different by a factor of 30). where N↑ and N↓ are the counts of coincidences for 4. Detectors of Electrons and Protons— different directions of neutron beam polarization. Determination of v/c Value Equation (3) indicates that a relative accuracy of A- asymmetry measurements of 10–3 can be reached when The electron detector will have a size of 55 × 160 all values: the average cosine of the A correlation coef- mm2. We plan to use a Si(Li)-detector for registration of ficient cos(σ p ), the neutron beam polarization P, the electrons. This detector will be divided into 16 individ- n e v electron velocity and the experimental asymmetry X ual sections. Previously we performed a test of a Si(Li)- c detector with a diameter of 70 mm and a thickness of are measured or determined with a relative uncertainty 3.5 mm to measure the backscattering effect of elec- better than 10–3. trons from Si. A magnetic β-spectrometer was used for this purpose. Thanks to these measurements we can dis- 3. Determination of Average Cosine of A tinguish the tail of the 976 keV line with total area Correlation Coefficient cos(σ , p ) 12 %. In our measurement with the correlation spec- n e trometer the effect of electron backscattering will be even more suppressed because the backscattered elec- The uncertainty of determination of the average trons will be reflected again from the magnetic plug cosine cos(σ p ) depends on the homogeneity of the n e with probability about 80 %. Those backscattered elec- magnetic fields in the decay region and in the region of trons that do go through the magnetic plug, will be reg- H ∆H − the magnetic plug. For 0 =≈×0.4 and 2 10 3 istered by the proton detector. This fraction will be HH m about 2.5 % only. We can reject these events due to a d cos(σ ,p ) we have:ne=×1.3 10−4 . Thus the average signal of prompt coincidence from both detectors. cos(σ ,p ) ne Therefore the energy resolution of the electron detector cosine of the A correlation coefficient can be deter- will be high enough. The energy calibration of this mined with very high accuracy. It can be attained by the detector can be done by means of a set of conversion magnetic collimation method, which is independent of sources. β the position of neutron -decay. The proton detector will have a size of 75 × 200 mm2, i.e., 20 mm broader and 40 mm higher than the

385 Volume 110, Number 4, July-August 2005 Journal of Research of the National Institute of Standards and Technology size of the electron detector. This is done to insure 7. Statistical Accuracy of the Experiment 100 % collection of protons. The proton detector will and the Random Coincidence have 16 sections as does the electron detector. This will Background help us to suppress the background of random coincidences and is necessary in any case to reduce the capac- The statistical accuracy of the experimental asymme- itance of the detector. We are going to use surface bar- 1 try ∆X is defined by , therefore 108 events have to rier detectors made from pure silicon. As a possible N alternative to this we are considering the use of be registered to obtain the accuracy ∆X ≈ 10–4. microchannel plates. The neutron flux density (Φ) at the exit of the polarized cold neutron beam of PSI fundamental facilities “Funspin” is about 2 × 108 cm–2s–1. The neutron flux 5. Precise Collimation of Neutron Beam density after the collimator will be one order of magni- Φ × 8 –2 –1 tude less ( C = 0.15 10 cm s ). We expect that the The task of precisely collimating the neutron beam is counting rate of neutron β-decay event (in coincidence) very important for our experimental scheme because will be about 110 s–1. Therefore 108 events will be col- the neutron beam passes close to the detectors. lected during two weeks. In this estimation it was The most appropriate collimator material is 6Li, assumed that the background of random coincidences is which practically does not produce any γ rays after neu- considerably less than the signal. tron capture. Because of its chemical activity, 6Li is The estimation of the background of the random used in form of 6LiF with some ingredients added to coincidence can be done using the background count- reduce the brittleness of LiF-ceramics. The small angle ing rate of the electron and proton detectors in similar scattering effect from this material was studied to experiments [6, 7]. The background of the random obtain initial data for the design of precise collimators. coincidences with a resolution time of 10 µs should be The experiment was carried out at the PNPI reflec- about (2 to 20) s–1. tometer of the WWR-M reactor. Using the data of the In order to obtain an exact answer concerning the small angle scattering process that was obtained, a background conditions, the test experiment with the multi-diaphragm collimator has been designed. The collimated neutron beam and the electron and proton distribution of neutron intensity at a distance of 4 m detectors is required. from the collimator was calculated. It was shown that the suppression factor of intensity reaches eight orders of magnitude at a distance of 2 cm to 3 cm from the 8. Conclusion edge of the beam. This result must be checked experi- mentally to be sure that satisfactory background condi- The above-mentioned considerations show that the tions can be reached. proposed experiment could reach an accuracy of A- asymmetry measurement at the level of (1–2) × 10–3. As a first step of the realization of the experiment, test 6. The Measurements of Neutron Beam measurements of background should be carried out. Polarization With an Accuracy Additionally, it should be mentioned that this spec- (1–2) × 10–3 trometer allows one to measure the B and a correlation coefficients as well. To do that, it is necessary to switch The possibility of precise measurements of neutron off the accelerating electric field in the spectrometer beam polarization has been studied in our detailed and to measure the longitudinal proton momentum by experiments [4, 5]. The scheme of the proposed method means of the time-of-flight method. Another way is to consists of two analyzers and two double flippers. use the electrostatic spectrometer in front of the proton The comparison of the two analyzers and two flip- detector. pers method with the method of 3He filters was done in our work [5]. It was shown that the correction for the supermirror method is 0.152(18) %. Thus, usage of the 9. References mentioned methods of neutron beam polarization determination can give measurement accuracies of 10–3 and [1] H. Abele, E. Barberio, D. Dubbers et al., Eur. Phys. J. C33, 1-8 better. (2004). [2] A. Serebrov, V. Varlamov, A. Kharitonov et al., this Special Issue.

386 Volume 110, Number 4, July-August 2005 Journal of Research of the National Institute of Standards and Technology

[3] H. Abele, M. Astruc, S. Baesler et al., Phys. Rev. Lett. 88, 211801 (2002). [4] A. Serebrov, A. Aldushchenkov, M. Lasakov et al., NIM A 357, 503-510 (1995). [5] O. Zimmer, P. Hautle, W. Heil et al., NIM A 440, 764-771 (2000). [6] B. G. Yerozolimsky et al., Phys. Lett. B412, 240 (1997). [7] H. Abele et al., Phys. Lett. B407, 212 (1997).

387