Search for Variations of Po-213 Half-Life
E.N.Alexeyev, Yu.M.Gavrilyuk, A.M.Gangapshev, A.M.Gezhaev, V.V.Kazalov, V.V.Kuzminov, S.I.Panasenko, S.S.Ratkevich
SSearch for Variations of
Po Half-Life
E.N. Alexeyev, Yu.M. Gavrilyuk, A.M. Gangapshev, A.M. Gezhaev, V.V. Kazalov, and V.V. Kuzminov
Baksan Neutrino Observatory INR RAS, Neitrino 361609, Russia
S.I. Panasenko and S.S. Ratkevich
V.N. Karazin Kharkiv National University, Kharkiv 61022, Ukraine
A device with the parent
Th source was constructed to search for variations of the daughter
Po half-life ( T / = 4 . µ s). A solar-daily variation with amplitude A So = (5 . ± . × − ,a lunar-daily variation with amplitude A L = (4 . ± . × − , and a sidereal-daily variationwith amplitude A S = (4 . ± . × − were found upon proceeding the data series over a622-day interval (from July 2015 to March 2017). The Po half-life mean value is found to be T / = 3 . ± . µ s. The obtained half-life is in good agreement with some of the literaturevalues obtained with great accuracy. Keywords : half-life,
Po nucleus, daily and annual variations
I. INTRODUCTION
Experimental studies of
Po half-life time depen-dence ( τ ) [1–3] have been carried out at the Baksan Neu-trino Observatory of the Institute for Nuclear Researchof the Russian Academy of Sciences since 2008. Unlikestudies on determining the half-life by the results of anal-ysis of time dependence of the explored isotope activity,decay curves analyzed at the Observatory are plottedbased on a set of data on lifetimes of separate nuclei ofthe Po isotope. In order to determine this parameter,the delays between the moment of a nucleus production(a beta electron from
Bi decay + a gamma-quantum)and its decay (an alpha particle from the
Po decay)are measured. The measurements are performed at TAU-2, a low-background facility placed in the undergroundlow-background laboratory NLGZ-4900 at the depth of4900 m.w.e. (973 days) and at TAU-1 of the undergroundlow-background KAPRIZ laboratory at the depth of 1000m.w.e. (354 days). The time interval of measurements atTAU-1 corresponds to the end of the measuring intervalat TAU-2. Further, time series of τ with different tem-poral steps are analyzed. According to data obtained atTAU-2, the averaged value of the Po half-life is τ =163 . ± . µ s. The annual variation with amplitude A = (9 . ± . × − , the solar-daily variation with am-plitude A So = (7 . ± . × − , the lunar-daily variationwith amplitude A L = (6 . ± . × − , and the sidereal-daily variation with amplitude A S = (7 . ± . × − are detected in the series of τ values.It was found that τ -amplitude maxima are observed atthe moments when the maximum projection is reachedby the Earths surface point velocity vector directed atthe explored source of possible variations (the Sun, theMoon, or an unidentified stellar object). Basically, itwould be possible to explain the origin of solar and lu-nar variations by these objects influence on performancesof measuring facilities through cyclical geophysical andclimatic disturbances (tidal waves, meteorological fac-tors, magnetic field, etc.) caused by them on the Earth. However, no reasonable fundamental process capable oftransforming these disturbances into variations of timeparameters of measuring facilities in the required phasehas been yet discovered.At the same time, the sidereal-daily variation of τ ob-served, if this is not an instrumental effect, may identifythe presence of a real unknown physical phenomenon in-fluencing the parameter under study. Two checks havebeen performed to verify the reliability of the results.The dependencies of the amplitude and phase of the ob-served sidereal-daily wave on the choice of starting pointof the analyzed time series were verified. The start wasshifted by 91 days and 182 days. The previous series sec-tion was not considered in the analysis. As expected forthis type of variation, waves were obtained with ampli-tudes coinciding with the initial values within the error,but with phases shifted by 6 and 12 hours relative to theinitial phases. The second verification was performed byplotting the daily data set in an anti-sidereal time (nonex-istent periodicity). The day duration in the anti-siderealtime was increased relative to solar day duration by aninterval shortening the day duration in sidereal time. Insuch a data set, within the statistical error ± . × − ,the wave was absent.Measurements were carried out at the TAU-1 facil-ity to confirm the non-random character of the observedvariations of the τ temporal series. Data provided byTAU-1 revealed a solar-daily variation with amplitude A So = (17 ± × − , a lunar-daily variation with am-plitude A L = (8 ± ± − , and a sidereal- daily variationwith amplitude A S = (11 ± × − . Searching for an-nual variations, initial data for a half year were summedand then sequentially shifted to increase the statisticalreliability of the decay curve. As a result, a τ series ofno more than six-month duration was obtained from theannual data set. Hence, it was impossible to identify theannual periodicity with adequate accuracy, so it was notexplored. It is clear that a substantial improvement inthe statistical error can be achieved by a substantial in-crease in the data-taking rate. However, in the case of a r X i v : . [ nu c l - e x ] J un FIG. 1. Decay schemes of
Bi and
Po. the
Po isotope, the speedup of data-taking induces aquadratic increase in the share of random coincidences,up to ∼
1% at a rate of 12 s − . This is caused by ahigh aggregate activity of all daughter Ra isotopesand the relatively long half-life of
Po ( ∼ . µ s).Therefore, without augmentation of the relative contri-bution of the random coincidence background, the in-crease in the statistics accumulation rate for Po canbe achieved only by the increase in the number of inde-pendent setups. This might appear technologically infea-sible. Another alternative is to use a pair of radioactiveisotopes having a similar decay scheme, but essentiallyshorter half-life of the daughter isotope. The followingisotopes might be used as such pairs:
Bi ( T / = 46min) → Po (weighted average of T / = 3 . µ s) [4]that are daughter products in the series of decays Th( T / = 7340 years) from the series of Np decays [5].In this paper, the first results obtained from using thefacility with the specified source are presented.
II. THE FACILITY DESCRIPTION
The construction of the TAU-3 facility with a
Thsource is similar to TAU-1 and TAU-2 [1]. It comprises ascintillation detector D1, a plastic scintillator (PS), whichis made of two disks d = 18 mm and h = 1 mm glued to-gether. The radiation source Th ( T / = 7340 years)positioned between the disks is the parent isotope for Po. The test sample is manufactured at the KhlopinRadium Institute (St. Petersburg). The source is pre-cipitated from Th(NO ) salt solution on the surface ofa LAVSAN film with h = 2 . µ m and covered by thesame film pasted along the edge by the epoxy resin. Theassembly is placed at the bottom of a case made of VM-2000 reflecting film open from one end. The case is putinside a stainless steel rectangular case 9 × ×
140 mm,thickness 0.5 mm. The open end of the case is connectedwith the bottom of a 2.5-mm stainless steel cylinder with d = 44 mm, and h = 160 mm. Inside the cylinder, thereis a high-speed FEU-87 photomultiplier monitoring PS. The signal is taken from the FEU anode load through thematching circuit and supplied via the cable (50 Ohm) tothe first entry of the registering detector. Detector D1is placed in the 15-cm Pb protective layer in a gap with h = 10 mm between two scintillation detectors NaI(Tl)150 ×
150 mm (detector D2) in a low-background box ofthe NLGZ-4900 underground low-background laboratory[6]. Signals from the anodes of two photomultipliers ofthe D2 detector are amplified by charge-sensitive pream-plifiers, summed, and supplied to the second, startingentry of the registering detector. The registering facilitycomprises a LA-n10-12 PCI digital oscilloscope (DO) in-tegrated with a PC that is registering in online mode thewaveform of pulses arriving from D1 and D2. The fre-quency of pulse digitization in DO is chosen as 100 MHz.The reading and recording is started by a pulse in theD2 channel. The record frame is 2048 temporal channels(10 ns per channel), including 256 channels of prehistoryand 1792 channels of history. In Fig. 1, the decays of
Bi and
Po isotopes [8] are presented schematically.From Fig. 1 a it follows that 66% of β decays of Bi aretransitions to the ground level, and 31% to the excitedlevel with an energy of 440 keV. The decay of this level isaccompanied by a γ quantum emission (26% per decay).The isotope of Po decays in 100% of cases with emis-sion of an α particle with an energy of 8537 keV. If thedevice registers all three particles released by the decayof the pair of isotopes, it is the event with three pulses.In this event, pulses coming from the γ quantum and β particle coincide instantaneously, and the pulse from the α is delayed. In Fig. 2, one of the events (frames) storedby DO in the PC memory is displayed as an example.The pulse on the upper beam ( ) is a γ quantum, thefirst pulse train on the lower beam ( ) corresponds toa β particle, and the second one to an α particle. Theobserved triple coincidences considerably reduce the con-tribution of background events accompanying decays ofthe remaining isotopes in the chain of decays of Thto the total counting rate of the facility. The activity of
Th is ∼
80 Bq. Alongside the main isotope there aresmall amounts of extraneous radioactive impurities in the
FIG. 2. An example of
Bi–
Po pair decay event storedby DO in PC memory: ( ) upper beam, a pulse from D2detector ( γ quantum), ( ) lower beam, pulses from particle(start), and particle (stop) in D1 detector. specimen. The rate of the event recording started by DOby D2 pulses with amplitudes of 380500 keV was ∼ − . The rate of recording useful events with parametersof all pulses corresponding to Po decay was ∼
18 s − .Following from Fig. 2, signals from and particles aretrains of short subpulses with total duration of up to ∼ µ s, decreasing exponentially in frequency and amplitude.The trains can overlap at small delays between particles;therefore, the processing program should consider rela-tions between the amplitudes of the first and subsequentsubpulses in a train in order to unambiguously separatethe delayed ( β ⊗ α ) coincidences.The delays between pulses in channel D1 are deter-mined by the results of processing the recorded oscillo-grams, and a decay curve of daughter isotope Po isplotted for the chosen time interval. The half-life de-termination is based on this curve. The sequential timeseries of this magnitude is plotted.
III. MEASUREMENT RESULTS
Continuous measurements started at TAU-3 on July9, 2015. The statistics for 622 days (March 2017) areprocessed. In Fig. 3, the decay curve of the
Po isotopeis given. The value of τ was obtained approximating thedecay curve by function F ( t ) = A × exp [ − ln (2) t/τ ] + b using the minimum χ test in the delay interval of 0.5-13.0 µ s. It was found that τ = 3 . ± . µ s.The primary data-consistent summation method wasused to find possible periodic dependencies. This is themethod of the interior moving average: to find harmonicsin a data series an interval is chosen about 0.5 of the ex-pected period, and the required parameter is determinedfor this interval; then the interval is shifted by one stepand the procedure repeats.In the studies of daily variations of the Po half-lifedepending on solar, sidereal, and lunar time, the length
FIG. 3. Decay curve for
Po plotted by the data fromTAU-3 device obtained over 622 days.FIG. 4. Dependence of
Po half-life on the time of so-lar day obtained by the method of interior moving average(triangles). Approximation by function τ ( t ) = τ [1 + 3 . × − sin { π/ t − } ] (red curve). Restored dependence τ ( t ) = τ [1 + 5 . × − sin { π/ t − } ] (blue dot-dashedcurve). of the respective day was divided into 24 hours. Thelength of a sidereal and of a lunar day in the standardsolar time is 23 hours 56 minutes 4.09 s and 24 hours50 minutes 28.2 s, respectively. A period of 12 hours waschosen as an interval of averaging. The analysis of eventswas made as follows. We selected the events registered inthe interval of 0-12 hours for the entire period study anddetermined the half-life values. After that, the intervalwas shifted by one hour and the procedure repeated. Theresults of the search of the daily variation in solar timeare given in Fig. 4. Here, the result of approximationof the daily half-life dependence by the function τ ( t ) = τ [1 + Asin { ω ( t + φ ) } ] (red curve) is displayed, where τ is the mean half-life; ω = 2 π/
24 h − ; A = 3 . × − isthe amplitude; φ = − Po half-life is welldescribed by a sinusoidal function. The period found is24 hours and the relative amplitude is 0.00034 half-lives.
FIG. 5. Dependence of
Po half-life on the time of side-real day obtained by the method of interior moving average(triangles). Approximation by function τ ( t ) = τ [1 + 2 . × − sin { (2 π/ t − } ] (red curve). Restored dependence τ ( t ) = τ [1 + 4 . × − sin { π/ t − } (blue dot-dashedcurve). It is easy to show that the initial periodic dependenceof time data has the same period (24 h), the amplitudeis higher by the factor of π/ . ×
24 = 6 h). The amplitudeof the initial daily periodic dependence obtained fromthese data in solar time is A So = (5 . ± . × − (bluedot-dashed curve).In Fig. 5, the results of the search for a siderealdaily variation of the Po half-life are displayed. Theexperimental data are approximated by curve τ ( t ) = τ o [1 + Asin { ω ( t + φ ) } ] (red curve) with parameters A =2 . × − is amplitude; φ = −
19 h is the phase shift ofthe curve initial point relative to 0 hours.The analysis of the restored initial dependence similarto the analysis made for the solar-daily wave shows thepresence of a sidereal-daily wave with relative amplitude A S = (4 . ± . × − (blue dot-dashed curve).In Fig. 6, the results of search for a lunar-daily vari-ation of the Po half-life are given. The analysis ofthe restored initial dependence like the analysis made forthe solar-daily wave shows the presence of a lunar-dailywave with relative amplitude A S = (4 . ± . × − (blue dot-dashed curve).In Fig. 7, the time dependence of τ obtained fromthe decay curve for a weekly data set is presented. Itis shown that τ increases with time, and that for adata set collected over 127 days, τ = (3 . ± . µ s; for 320 days, τ = (3 . ± . µ s; for 422days τ = (3 . ± . µ s, and for 622 days τ =(3 . ± . µ s. The causes of such behavior of the τ parameter are not clear yet. It could be both an in-strumental effect, for example, equipment ageing, and anunknown real physical effect. It seems impossible to fore-cast further curve tendency, and only further consistentmeasurements can possibly provide the solution of thisproblem. The presence of a pulse surge of data withinthe time interval comparable to a year in the series of FIG. 6. Dependence of
Po half-life on the time of lu-nar day obtained by the method of interior moving average(triangles). Approximation by function τ ( t ) = τ [1 + 3 . × − sin { π/ t − } ] (red curve). Restored dependence τ ( t ) = τ [1 + 4 . × − sin { π/ t − } ] (blue dot-dashedcurve).FIG. 7. Time dependence of τ obtained from the decay curvefor the weekly data set (start of measurements: July 9, 2015). weekly data hinders using the method of moving inter-nal average for studies of the half-life annual variation.Therefore, in order to get a better understanding of theannual variation, we have checked a supposition that thehalf-life annual variation detected in the series of weeklydata on the half-life of the Po isotope [3] continueswith the same amplitude and phase in the series of dataon
Po. The data normalized to unity for these iso-topes in continuous time are presented in Fig. 8. Thedata normalization for
Po was made using the τ meanvalue for 320 days. It is shown that the annual variationof the Po data with the same amplitude and phase asthe data on
Po is not excluded. The repeated devia-tion, after its shape is specified, can be removed from theseries of data on
Po to study the remainder for annualvariations.
FIG. 8. Superimposed time series of weekly mean normal-ized quantities of
Po half-life (0-140th week, 973 days) and
Po (146-234 th week, 622 days). The sinusoidal functionshows an approximation of the annual variation of Po half-life; it is extrapolated over the entire interval of observations.FIG. 9. Time dependence of
Po half-life measured fromthe beginning of the 38.73-day interval by the method of inte-rior moving average (triangles) with a one-day step. Approx-imation by function τ ( t ) = τ [1 + 6 . × − sin { π/ . t − . } ] (red curve). IV. DISCUSSION OF RESULTS
The above-presented results of the
Po isotope-decay-constant monitoring for the period of July 2015March 2017 show that this parameter was subject tosolar-daily, sidereal-daily, and lunar-daily variations withamplitudes A So = (5 . ± . × − , A S = (4 . ± . × − and A L = (4 . ± . × − . Within the error,these values coincide with the amplitudes of correspond-ing variations detected in the series of half-lives of Po.The search for annual variations in
Po data is com-plicated by the occurrence of aperiodic single-sided de-viations of half-lives from the average values in the timeseries. The process became visually noticeable in thesection of data recorded over May-June 2016. The effectmay be caused by both equipment ageing and unknown physical factors. To find the answer we need to continuemeasurements.A component with frequency of 9.43 yr − (period of38.73 days) and the maximum power was detected overthe period of June 1996–July 2001 in the analysis ofthe power spectrum of frequency components composingthe series of counting rates ( ∼ / τ ( t ) = τ [1 + Asin { ω ( t + φ ) } ], where τ is the mean half-life; t istime [days]; ω = 2 π/ .
73 day − ; A = (6 . ± . × − is amplitude; φ = − . A = (10 . ± . × − was obtained fromthe approximation by multiplying by π/
2. In order toverify the result stability, a similar procedure was per-formed with the data of the TAU-2 facility accumulatedfor 590-day measurements with the
Po isotope. Therestored wave amplitude was A = (10 . ± . × − .The search for a wave with the frequency of 10 yr − was carried out for verification. Within the statisticalerror of ± . × − , no variation with this frequencywas detected. Thus, we can conclude that the variationwith the frequency of 9.43 yr − is of global character (atleast, for the underground devices), though does not co-incide with known natural rhythms. In [7], the authorsconsider the possibility that the Suns core can feature asimilar rhythm, and study possible mechanisms of thesevariations. V. CONCLUSIONS
In this study, to search for variations of
Po half-life ( T / = 3 . µ s), a device with Th isotope as aparent source was constructed. A solar-daily variationwith amplitude A So = (5 . ± . × − , a lunar-dailyvariation with amplitude A L = (4 . ± . × − , and asidereal-daily variation with amplitude A S = (4 . ± . × − were discovered upon processing the data series forthe period from July 2015 to March 2017 (622 days). The Po half-life value averaged over 662 days is found tobe T / = 3 . ± . µ s. This is consistent with theresult ( T / = 3 . ± . µ s) obtained by means ofan ion-implanted planar Si detector for alpha and betaparticles emitted from weak Ac sources in work [8].
ACKNOWLEDGMENTS
The study was supported by the High Energy Physicsand Neutrino Astrophysics Program of the Presidium ofthe Russian Academy of Sciences. [1] E.N. Alexeyev, V.V. Alekseenko, Ju.M. Gavriljuk,A.M. Gangapshev, A.M. Gezhaev, V.V. Kazalov,V.V. Kuzminov, S.I. Panasenko, S.S. Ratkevich, andS.P. Yakimenko, “Experimental test of the time stabil-ity of the half-life of alpha-decay
Po nuclei,” Astropart.Phys. , 23 (2013).[2] E.N. Alexeyev, Yu.M. Gavriljuk, A.M. Gangapshev, V.V.Kazalov, V.V. Kuzminov, S.I. Panasenko, and S.S. Ratke-vich, “Sources of the systematic errors in measurementsof Po decay half-life time variations at the Baksan deepunderground experiments”, Physics of Particles and Nu-clei, , 157 (2015).[3] E.N. Alexeyev, Yu.M. Gavriljuk, A.M. Gangapshev, V.V.Kazalov, V.V. Kuzminov, S.I. Panasenko, and S.S. Ratke-vich, “Results of a search for daily and annual variationsof the Po half-life at the two year observation period”,Physics of Particles and Nuclei, , 986 (2016).[4] G. Audi, F.G. Kondev, M. Wang, B. Pfeiffer, X. Sun,J. Blachot, M. MacCormick, “The Nubase2012 evaluationof nuclear properties”, Chin.Phys. C , 1157 (2012). [5] Reference book edited by I. K. Kikoin “Tables of physicalmagnitudes”, (Atomizdat, Moscow, 1976) [in Russian].[6] Yu.M. Gavriljuk, A.M. Gangapshev, A.M. Gezhaev,V.V. Kazalov, V.V. Kuzminov, S.I. Panasenko,S.S. Ratkevich, A.A. Smolnikov, and S.P. Yakimenko,“Working characteristics of the New Low-BackgroundLaboratory (DULB-4900)”, Nucl. Instr. Methods in Phys.Res. A , 576 (2013).[7] P.A. Sturrock, E. Fischbach, and J.D. Scargle, “Compar-ative Analyses of Brookhaven National Laboratory Nu-clear Decay Measurements and Super-Kamiokande SolarNeutrino Measurements: Neutrinos and Neutrino-InducedBeta-Decays as Probes of the Deep Solar Interior”, SolarPhysics Fr,
At,
Bi,
Po and
Pb fromthe
Ac decay series”,Applied Radiationand Isotopes77