SEARCH

SEARCH BY CITATION

Keywords:

  • β−ν angular correlation;
  • low-energy precision measurements;
  • weak interaction;
  • paul trap

Abstract

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The experimental achievements and the current program with the LPCTrap device installed at the LIRAT beam line of the SPIRAL1-GANIL facility are presented. The device is dedicated to the study of the weak interaction at low energy by means of precise measurements of the β−ν angular correlation parameter. Technical aspects as well as the main results are reviewed. The future program with new available beams is briefly discussed.


The β−ν angular correlation

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The structure of the weak interaction remains an important topic in the area of fundamental interaction physics. In particular, at low energies, there are nowadays ambitious experimental programs to search for “exotic” currents beyond the V–A theory [1, 2]. Such studies have recently made significant progress with the advent of improved trapping techniques [3, 4]. These sophisticated setups allow the production of β-decay sources almost at rest and confined in small volumes that can be surrounded by suitable detectors [5, 6]. We describe here the achievements made with the LPCTrap setup installed at GANIL1. The experimental program is dedicated to precision measurements of the β−ν angular correlation parameter.

From a theoretical point of view, the generalization of Fermi's theory leads to consideration of five different Lorentz invariant contributions in the transition amplitude describing nuclear beta decay, which are referred to as scalar (S), vector (V), tensor (T), axial-vector (A) and pseudoscalar (P). The search for S and T exotic contributions can be performed through a precise measurement of the β−ν angular correlation parameter, aβν. For allowed transitions and nonoriented nuclei, this parameter can be directly inferred from an events distribution linked to the angular correlation between the leptons [7]. Since neutrinos are difficult to detect, a sensitive observable for a β−ν angular correlation measurement is a kinematic parameter of the recoiling daughter nucleus. If the β particles and the recoil ions are detected in coincidence and the relative time of flight (ToF) is measured between the two particles, the expected distribution of events in the electron energy and recoil ion time of flight, t, is for all solid angles detected:

  • display math(1)

where Te, Ee and r(t) denote the kinetic and total energies of the β particle and the recoil ion momentum, respectively. Q is the energy available in the transition, K is a constant and FZ, Ee) is the Fermi function (β and β+ decays). The parameter aβν and the Fierz interference term b depend on the coupling constants, Ci and Ci′ (i = S, V, T, A), associated to the different contributions [7]:

  • display math(2)
  • display math(3)

where

  • display math(4)

MF and MGT are the Fermi (F) and Gamow–Teller (GT) nuclear matrix elements, α is the fine structure constant, Z is the atomic number of the daughter nucleus and m and pe are the mass and momentum of the β particle, respectively. The P contribution is neglected in a nonrelativistic description of the nucleons. The existence of both coupling constants, Ci and Ci′, is related to the transformation properties under parity. The Standard Model (SM) assumes maximal parity violation (|Ci| = |Ci|), time reversal invariance (Ci, Ci′ real) and pure vector and axial-vector interactions (V–A theory). The β−ν angular correlation parameter is then given by:

  • display math(5)

where inline image is the mixing ratio of the transition, leading to aβν = 1 for a pure F transition and aβν = –1/3 for a pure GT transition.

Under the conditions of the present experiment, the parameter that is determined experimentally is actually:

  • display math(6)

where the brackets < > mean a weighted average over the measured part of the β spectrum.

For mirror transitions, the measurement of aβν also allows for a precise determination of the mixing ratio ρ. If the half-life, the branching ratio (BR) and the masses involved in the transition are well known, this ratio can be used to deduce the first element of the Cabbibo–Kobayashi–Maskawa (CKM) quark mixing matrix, Vud [8]:

  • display math(7)

where Ft1/2 is the corrected ft-value of the transition, GF is the weak interaction fundamental constant, ΔR is a common radiative correction and fA (fV) is the rate statistical function computed for the GT (F) component. Many interesting mirror transitions exist [9], which could form an additional database to test the conserved vector current (CVC) hypothesis in nuclei and to determine Vud.

The LPCTrap setup

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The experimental setup is installed at the low-energy beam line LIRAT2 of the SPIRAL1-GANIL3 facility. Technical details of the setup are described elsewhere [5, 10-12]. The central element of the device is a transparent Paul trap (Fig. 1) constructed to confine singly charged radioactive ions, almost at rest, in a small volume, allowing the detection in coincidence of the β particles and of the recoiling ions. The trap consists of 6 concentric rings with an open geometry, allowing easy injection and extraction as well as an efficient detection of the decay products.

image

Figure 1. Picture of the transparent Paul trap of the LPCTrap setup.

Download figure to PowerPoint

Upstream of the trap, the installation is equipped with a radio-frequency quadrupole cooler-buncher (RFQCB) for the preparation of the beam by reducing the emittance and the production of ion bunches. The low-energy radioactive beam is provided by the ECR source of the SPIRAL facility with a typical energy dispersion of about 20 eV at 10 keV total kinetic energy. The RFQCB is connected to the Paul trap by a short transport line with dedicated beam optics and diagnostics. The offline tuning of the ensemble is made possible by the coupling of an ion source at the entrance of the setup. There, a Faraday cup and a retractable silicon detector provide efficient beam diagnostics. A layout of the equipment is shown in Fig. 2.

image

Figure 2. Scheme of the full LPCTrap setup.

Download figure to PowerPoint

The incident beam is first decelerated below a few tens eV by the H.V. applied to the RFQCB platform. This enables the cooling of the ions by collisions with a buffer gas (H2 or 4He) at a pressure of the order of 10−2 mbar. Beam bunches are produced near the exit of the quadrupole at a repetition rate of the order of 10 Hz (cycle). A first pulse cavity is used to transport the bunches at 1 keV downstream from the RFQCB. A second pulse cavity reduces the kinetic energy of the ions at 100 eV for an efficient injection into the Paul trap. There, the ions are once again cooled by elastic collisions with an inert gas injected at very low pressure (10−6–10−5 mbar). At equilibrium, the thermalized ions have energies of about 0.1 eV and the diameter of the ion cloud located at the center of the trap is of the order of 2.5 mm (see Fig. 7 in the next section).

Two different detection setups, shown in Fig. 3, have been used in the experiments. In both setups, a telescope and a microchannel plate position sensitive detector (MCPPSD) detect the β particles and the recoil ions, respectively. In the first setup, both detectors are located 10 cm away from the center of the trap in a back-to-back configuration. The β telescope is made of a 60×60 mm2 300-μm thick double-sided silicon strip detector (DSSSD) for position readout, followed by a 7-cm thick plastic scintillator coupled to a photomultiplier to measure the β particle kinetic energy and to deliver the start signal for the recoil ion time-of-flight measurement. The stop signal is delivered by the MCPPSD whose performances are detailed in Ref. [13]. For each coincidence event, the positions of both particles, the recoil ion ToF, the β particle kinetic energy, the time stamp of the decay within the cycle and the RF phase of the trap are recorded. This set of measured parameters is useful to control systematic effects and to check the consistency of the results.

image

Figure 3. Top view of the detection chamber of LPCTrap. Left panel: First detection setup used in the commissioning phase. Right panel: Improved detection setup allowing the separation of the charge states of the recoil ions.

Download figure to PowerPoint

In the second setup implemented in 2010 (Fig. 3, right panel), a spectrometer has been added to separate the charge states of the recoil ions. The ions emitted towards the spectrometer are accelerated by a –2 kV potential after they cross the collimator located in front of a 50-cm long free-flight tube. An electrostatic lens set at −250 V allows the collection of all the ions on the MCPPSD located at the end of the tube. The front plate of the detector is set at –4 kV to ensure a maximum detection efficiency for all charge states, independently of the ion kinetic energy. This spectrometer makes LPCTrap a unique and unprecedented setup for the measurement of charge-state distributions of ions associated with the β decay of singly charged radioactive ions.

Measurements in the decay of 6He

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The first nucleus of interest that has been considered is 6He, which decays via a pure GT transition. For such transitions, the experiment described here was the first performed after decades following the last measurements of the beta-neutrino angular correlation in the β decay of 6He [14-16] and 23Ne [17, 18]. Another new experiment, measuring the β−α−α correlation from trapped 8Li ions, has recently reported the first results [19]. Among these experiments, only one measurement in the decay of 6He [14] was performed with a relative precision at the level of 1%, yielding aβν = −0.3308(30) after inclusion of radiative and recoil-order currents corrections [20].

At GANIL, the 6He1+ beam is produced by the fragmentation of a primary 13C beam at 75 MeV/u on a graphite target coupled with an ECR source. The low-energy beam is delivered at 10 keV through the LIRAT beam line with a maximum intensity of about 1.5×108 pps. The resolving power (MM ∼ 250) of the dipole magnet located at the entrance of the low-energy beam line is not sufficient to eliminate the 10 nA of a stable beam containing mainly 12C2+, but this contamination is drastically reduced in the RFQ.

A first experiment was performed using the initial setup (Fig. 3, left panel) with a measurement cycle of 100 ms. This experiment is described in details in [21]. The ToF spectrum of the recoil ions is fitted to extract aβν. The analysis is based on the comparison between the experimental ToF spectrum and those obtained for two sets of realistic Monte Carlo (MC) simulations considering pure axial (aβν = –1/3) and pure tensor (aβν = +1/3) couplings. In a first step, the experimental data are calibrated and corrected for the identified sources of unwanted events (background). Then, aβν is deduced from a fit of the experimental ToF spectrum with a linear combination between the two sets of simulated decays obtained for axial and tensor couplings. Among the many instrumental effects, the detectors response function and geometry, the trap RF field influence, the ion cloud characteristics inside the trap, the shake-off ionization of the recoil ion, and the scattering of the β particles are implemented in the simulations.

The calculation of the expected value of the angular correlation parameter in the SM includes radiative corrections [20, 22]. We have used the formalism described by Glück [20], based on the work of Sirlin [23] to calculate, to first order in α and on an event-by-event basis, the change in the kinematics due to the virtual and real photon emission during the decay process. It turns out that such corrections are at the 1% level of the value of the correlation parameter.

The shake-off ionization of the recoil ion due to the sudden change of the electric charge of the nucleus following β decay is also taken into account. For the present analysis, the shake-off ionization probability of 6Li2+ ions has been calculated in the sudden approximation, and found to be 0.02334 + 0.00004ERI, where ERI is the ion recoil energy in keV. Since the maximum recoil energy is 1.4 keV, the energy-dependent term can be neglected. This ionization probability is in perfect agreement with previous calculations of Wauters and Vaeck [24]. In a second experiment performed with the improved setup of Fig. 3, the shake-off probability was experimentally measured and was found to be in very good agreement with the theoretical value [12].

The characteristics of the ion cloud inside the trap have a strong influence on the ToF spectrum. Simulations of the trapped ions trajectories in the Paul trap have thus been performed using the SIMION8 software package [25]. The geometry of the electrodes of the surrounding elements was included as well as the experimental RF trapping voltages. The collisions between the trapped ions and the H2 buffer gas molecules were described at the microscopic level using realistic interaction potentials [26]. The position and velocity distributions at thermal equilibrium as a function of the RF phase can be obtained from such simulations. Based on these calculations, the mean thermal energy of the ion cloud is expected to be kTsim = 0.09 eV. This is smaller than the value obtained in offline measurements, kTexp = 0.107(7) eV [27]. A small correction (scaling) factor was thus applied to the cloud characteristics in the simulations. Measurements performed at different ion cloud densities have shown that space-charge effects are negligible under the adopted measuring conditions. The response functions of the detectors dedicated to the detection of the electron and the recoil ion were simulated with particular emphasis on the influence of the backscattering of β particles on the detectors and on other structures inside the detection chamber.

An important source that affects the shape of the ToF spectrum is the backscattering of electrons since, in this case, the kinematics of the decay is biased. The ToF distribution associated with such events (“scattered”) is obtained by massive simulations using the GEANT4 software code.

False coincidences (“accidentals”), corresponding to the detection of uncorrelated particles on the β telescope and on the microchannel plate, also contribute to the ToF spectrum and introduce a flat distribution that is easily subtracted. It appears that most such accidentals events originate from (i) the decaying neutralized 6He species that are present inside the whole volume of the chamber and trigger the β telescope, and (ii) H2 molecules leaking from the RFQCB that trigger the MCPPSD detector.

For a small fraction of the 6He atoms decaying in the chamber volume, the recoil ion can be detected on the MCPPSD, in coincidence with its associated β particle. This constitutes another source of physical events labelled “out-trap” events. Their contribution to the ToF spectrum in the region of the fit can bias the measurement. They are included in the simulation, assuming a decay process with pure axial coupling.

The ToF spectrum obtained for validated coincidences is displayed in Fig. 4. The selected events are conditioned by: a 500 keV energy threshold on the energy deposited in the scintillator, a time within the trapping cycle between 25 and 95 ms (cloud at thermal equilibrium), and a valid reconstruction of the positions in both the DSSSD and the MCPPSD. The “scattered”, “out-trap”, and “accidentals” events represent, respectively, 4.5%, 2.6%, and 7.3% of the total number of events. These contributions and the ToF obtained for aβν = –1/3 are summed and compared to the experimental spectrum in Fig. 4.

image

Figure 4. Experimental time-of-flight spectrum (black line) compared to the simulated one (gray area) in the pure axial case, including the different simulated contributions of background, after normalization (see text for details).

Download figure to PowerPoint

By considering the β decay vertex as a point-like source at the center of the Paul trap, the recoil ion ToF and position can be used to determine the three components of the recoil ion momentum. In a similar way, the full momentum vector of the β particle can be deduced from the energy deposited in the β telescope and the position on the DSSSD. This provides the possibility to reconstruct the antineutrino invariant mass:

  • display math

This variable is useful to validate in a global manner the quality of the simulations with a small influence of the value of the correlation parameter. Fig. 5 shows the antineutrino invariant mass spectra obtained for the experimental and simulated events. The main peak is well reproduced by the simulations. The shape and position of this peak depend on all the inputs of the simulations (background, detector response functions, geometries, size of the ion cloud, trap RF field, etc.). The agreement between the experimental and simulated data provides an additional test of the inputs used in the analysis.

image

Figure 5. Antineutrino invariant mass spectra for the experimental data, the simulation in the pure axial case, and the simulation of background events. Vertical lines indicate the selection window used in the analysis to reduce background contribution.

Download figure to PowerPoint

After applying a window cut in the antineutrino invariant mass spectrum and after background subtraction, the experimental ToF spectrum is adjusted with a linear combination of the time-of-flight spectra simulated using pure axial and pure tensor couplings (Fig. 6, left panel). Three parameters were left free in the fit: the value of aβν, the total number of events, and the distance dMCPPSD between the MCPPSD detection plane and the center of the Paul trap. Such a distance cannot be measured with better accuracy than a fraction of a mm and is therefore a degree of freedom to be taken into account in the analysis. The range in ToF selected for the fit is indicated by the vertical lines. The experimental data have first been split randomly in four independent sets, and the corresponding ToF spectra were fitted by changing the upper limit of the fitting range. No significant dependency has been found. Contours of constant χ2 in the plane of parameters dMCPPSD and aβν are shown on the right panel in Fig. 6. The result from the fit leads to aβν = –0.3335 ±0.0073, and dMCPPSD = 100.255 ±0.011 mm. The nominal value for this distance is 100.0 mm with a positioning uncertainty of 0.5 mm. The minimum chi-square χ2min = 96.6 for 105 degrees of freedom corresponds to a P-value of 0.71, which indicates a very good agreement between the data and the fitted function. The quoted error is purely statistical.

image

Figure 6. Left panel: Fit of the experimental spectrum (upper panel). The range of the fitted function is indicated by vertical lines. The normalized residuals are plotted in the lower panel. Right panel: projection on the plane of parameters dMCPPSD and aβν of the computed contours for (χ2χ2min) values = 1, 2, and 3.

Download figure to PowerPoint

The contributions of the main sources of systematic uncertainty are listed in Table 1. The label “present data” in the column “Method” of Table 1 indicates that the parameters and their uncertainties were determined by fitting the experimental data with the MC simulation (assuming a pure axial coupling). In each case, it was verified that the sensitivities of these parameters to the value of aβν taken as input were negligible at the current level of precision. The associated uncertainties on aβν were then deduced from the changes in the aβν values resulting from the fit of the experimental ToF spectrum while varying the parameters in the MC simulation.

Table 1. Main sources of systematic error, systematic uncertainties, and impact on the error of aβν that are discussed in the text. The total uncertainty quoted corresponds to the sum of all effects taken into account in the analysis for which some of them are not discussed here
SourceUncertainty Δaβν (×10−3)Method
Cloud temperature6.5%6.8offline measurement
“accidentals” and “out trap”see text0.9present data
β Scattering10%1.9GEANT4
Total (see [21]) 7.5 

It has been found that the aβν value resulting from the fit strongly depends on the trapped ion cloud size and on the temperature used in the MC simulation (Fig. 7). This motivated an independent measurement of the ion temperature using an offline source of 6Li+ [27], performed under identical running conditions to those in the 6He+ experiment in terms of trap RF voltage, gas pressure in the trap chamber and number of trapped ions. A relative precision of 6.5% was obtained that constitutes the dominant contribution to the systematic error on the value of aβν.

image

Figure 7. Values of the correlation coefficient resulting from the fit procedure as a function of the ion cloud thermal energy, kT, and of the ion cloud square radius, r2, where r is the RMS of the spatial distribution in the Paul trap radial plane. The dashed and solid lines correspond, respectively, to the central value and the 1σ uncertainty of the offline temperature measurement.

Download figure to PowerPoint

The uncertainties due to “accidentals” and “out-trap” event subtractions are only statistical. They are limited by the statistics of experimental events that serve as normalization for the MC results. The uncertainty due to the “scattered” events was estimated by considering a 10% relative error on the β scattering yield provided by the GEANT4 simulations. This 10% relative error is based on the work of Hoedl [28], which compares a compilation of published electron scattering experimental data to several MC codes.

Combining all systematic errors quadratically, the final result is:

  • display math

where the first uncertainty is statistical and the second systematic.

Figure 8 shows the first result obtained with LPCTrap compared with other measurements of the β−ν angular correlation coefficient in pure GT transitions. The values from Carlson and from Allen et al. were obtained in 23Ne decay, the value of Li et al. was obtained in 8Li decay and the others in 6He decay. Our result and the one of Li et al. are the most accurate among the experiments performed using the detection of the recoil ions and β particles in coincidence. The measurement presented here, performed with a different and independent method, confirms the result of Johnson et al. It is to be recalled that the reduced chi-square χ2/ν = 0.92 for 105 degrees of freedom obtained in the present work corresponds to a P-value of 0.71. This is to be compared to a P-value of 0.055 for the Johnson et al. experiment, with a reduced chi-square χ2/ν = 1.69 for 13 degrees of freedom [20].

image

Figure 8. From top to bottom: aβν experimental values in pure GT transitions from [14, 17, 18] (left panel), [19], present work, [15], and [16] (right panel). The error bars show the quadratic sums of statistical and systematic uncertainties. The dashed lines indicate the value predicted by the SM.

Download figure to PowerPoint

The techniques used in the two experiments differ in a number of other aspects. First, the use of trapping techniques and the detection in coincidence of two decay products resulted in a larger signal to background ratio, by more than a factor of two compared to the Johnson et al. experiment. Furthermore, the measurement in an event-by-event mode of additional parameters (particle positions, energy of the β particle, RF phase, and time within the trapping cycle) allows a better control of possible systematic effects. With the efficiencies achieved for beam preparation and trapping and for the detection of coincidence events, the average counting rates in the present experiment was about 1 coincidence per second and about 100 single triggers per second. This is, respectively, three and one orders of magnitude lower than in the Johnson et al. experiment. Thus, possible systematic effects due to the counting rate are expected here to have a smaller impact. The dominant contribution to the uncertainty in the Johnson et al. experiment was due to random variations of recoil-energy spectra in a number of sequential data acquisitions, while in the present work, the precision limitation is mainly statistics. Both techniques are thus complementary.

From Eq. (6) and using the value <mc2/Ee> = 0.2 deduced from the β particle energy spectrum [21], the result from the first experiment with LPCTrap corresponds to a limit on the tensor coupling

  • display math

assuming CT = CT′.

A second experiment has been performed with the setup involving the recoil ion spectrometer (Fig. 3, right panel) [12]. A major improvement in the efficiency was reached resulting from several modifications of the setup. First, the injection inside the RFQ was optimized as well as the transmission between the RFQ and the trap. This led to the following efficiencies for a measurement cycle of 200 ms: 0.6% through the RFQ, 40% downstream up to the trap, 20% inside the trap. With an initial intensity of 108 pps, this corresponds to about 104 ions/bunch inside the Paul trap. The efficiency of detection was also increased by modifying the geometry of the setup. A total of about 1.2×106 true coincidences were collected during four days of data taking.

In addition, the understanding of the systematic effect associated with the cloud temperature has been improved by studying in further details the behavior of the leading edge of the ToF distribution [29]. The high statistic of the experiment also enables to improve the subtraction of background by carefully considering the different contributions of such events as a function of time inside the measurement cycle. In the second run, the ions were ejected from the trap after 150 ms trapping duration to detect “out-trap” events during the 50 ms left.

As already stated above, an experimental determination of the shake-off probability was made possible with a very high accuracy. Fig. 9 shows the experimental ToF distribution for the recoil ions detected in coincidence with the electron following the beta decay of 6He1+ with the background subtraction described previously. In the present case, only two charge states are possible (Fig. 9, left). The fit, which takes into account all the effects mentioned above, is in perfect agreement with the experimental data. The value obtained for the shake-off probability is:

  • display math
image

Figure 9. Left panel: Fit of the experimental spectrum to deduce the charge-state distribution. Right panel: Dominant source of systematic effects.

Download figure to PowerPoint

The main sources of systematic effects have been investigated (Fig. 9, right), and the total systematic uncertainty is very small compared to the statistical error. The experimental value of Pshake-off is in excellent agreement with the theoretical prediction based on the sudden approximation.

As far as the angular correlation parameter is concerned, a preliminary value has been estimated from another analysis: aβν = –0.3338(26)stat. The error on this value is dominated by the statistics used for the number of events in the simulation. The value is consistent with our previous result and the value obtained by Johnson et al.

The analysis of the experimental data is about to be finalized and the statistics gathered during the experiment should allow us to reach an unprecedented statistical accuracy of 0.0015.

Measurements in the decay of 35Ar

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The 35Ar nucleus decays through a mirror transition with a large Fermi fraction x = 92%. The basic parameters (T1/2, BR, masses) involved in the decay being well known [9], a value of aβν with sufficient precision can be determined in the framework of the SM. The precise measurement of aβν in this transition can then be interpreted either to constrain scalar currents, or to determine Vud. In a previous measurement, performed decades ago [17], the recoil ions after beta decay were detected in singles, leading to the value aβν = 0.97(14).

At the SPIRAL1-GANIL facility, the 35Ar1+ ions are produced by a primary 36Ar beam at 80 MeV A−1 impinging on a graphite target coupled to an ECR source. The radioactive beam is delivered to the LPCTrap at 10 keV with a typical intensity of 3.5×107 s−1 (∼ 5.5 pA), measured by the retractable silicon detector. The dipole magnet located at the entrance of the low-energy beam line is again not sufficient to eliminate the 40 pA of a stable beam containing mainly molecules that have not been fully identified.

The tuning of the RFQ has been adapted to the 35Ar1+ ions that are cooled with He buffer gas at a 1.6×10−2 mbar pressure. To prepare the experiments, tests were first carried out with different stable beams produced either by a surface ionization source or the ECR source of SPIRAL. Typical performances reached with the different beams are given in Table 2 for two values of the cycle (20 ms and 200 ms).

Table 2. Efficiencies of the three main sections of LPCTrap obtained with different stable and radioactive ions around mass 35, for two values of the cycle (20 ms and 200 ms)
  ε (RFQ)  ε (total)
SpeciesSource20 ms200 msε (transfer line)ε (trap)20 ms200 ms
39K1+Offline0.270.170.430.20.0230.015
36,40Ar1+ECR/SPIRAL0.320.120.220.150.010.0037
35Ar1+ECR/SPIRAL/0.150.250.10/0.0038

Due to space-charge effects, the transmission, and consequently the lifetimes of the ions in the buncher and in the Paul trap, can be limited by the number of ions per bunch. For the values given in Table 2, the beam current at the entrance of LPCTrap was adjusted in each case to keep this number below 107 ions per bunch in order not to saturate the RFQ. For a cycle of 200 ms, this number is reached with a typical beam intensity of 50 pA at the entrance of LPCTrap. A maximum of 2.5×105 ions can then be confined in the Paul trap. In the RFQ, the measured efficiencies are mainly defined by the losses at beam injection in the setup (the hole at the entrance has a 6 mm diameter) and by the losses due to charge exchange. A short cycle gives access to the first factor: roughly 70% of the beam is lost at 20 ms. The values obtained for a longer cycle enable determination of the lifetime of the ions in the buncher due to trapping losses. This is of the order of 100 ms for the Ar+ species and 250 ms for K+ ions, which is consistent with a lower charge exchange expected with alkali ions. In the measurement Paul trap, the lifetime of confined ions has also been determined and is close to 500 ms for all species, with buffer He gas being injected in the chamber at a very low pressure.

During the last run performed with 35Ar1+, an average of 2.5×104 radioactive ions were confined for each bunch injection, every 200 ms. Even if the total efficiency of LPCTrap is lower for the stable contaminant, the number of ions counted in the ion cloud monitor revealed that a total of 1.5×105 ions were actually confined in the Paul trap at each cycle. This does not jeopardize the experiment, but space-charge effects should be considered in the data analysis. The trapped ions reached their thermal equilibrium about 20 ms after injection in the He buffer gas at a pressure of 1.5×10−5 mbar. After 160 ms trapping duration dedicated to the correlation measurement, the ions were extracted towards the ion cloud monitor and “out-trap” events detected during the 40 ms left. The second detection setup involving the recoil ion spectrometer (Fig. 3, right panel) was used to measure the decay products and a total of 1.5×106 real coincidences were recorded. The ToF spectrum obtained during this run is shown in Fig. 10.

image

Figure 10. Experimental ToF spectrum obtained during the run with 35Ar, for about 1.5×106 coincidence events.

Download figure to PowerPoint

The different charge states of the 35Cl daughter ions are well separated by the recoil ion spectrometer. The charge-state distribution had already been estimated from the commissioning run [30]. The preliminary values are:

  • display math

The study of systematic effects is currently ongoing. Among these effects, the probability of charge-exchange collisions in the ambient gas, which was negligible in the case of 6Li in H2 [12], has to be estimated in the present case involving many electrons.

The number of neutral recoil 35Cl atoms, which are not detected by the MCPPSD, can be deduced from the comparison between the number of expected coincidences (including neutrals) and the effective number of measured coincidences. The number of expected coincidences is simply computed using the number of β particles detected in singles and the overall ion detection efficiency. This estimate leads to 72(10)% of neutral 35Cl recoils. Beyond the prototypical 6He1+ case, this heavier system can reveal the role of more subtle shake-off dynamics involving many electrons as, for instance, the Auger processes [31].

The number of coincidences measured during the last run corresponds to a statistical uncertainty of 0.002 on aβν. In the study of the systematic effects, particular attention will also be devoted to the trap RF influence on the recoil motion. In the 35Ar decay, the recoil kinetic energy reaches at most 452 eV, which is about a factor of 3 lower than in the case of 6He. The RF effect could then become a dominant source of the systematic uncertainty. Moreover, as the total number of ions in the trap nearly reaches the trap capacity, the space-charge effect has to be taken into account in the analysis. Finally, the second main branch of the decay with γ emission (BR = 1.23%) has also to be considered. Such systematic contributions should be properly managed and, assuming a systematic uncertainty of the same order as the statistical one, the final result would constitute the most precise value ever obtained in a β−ν angular correlation measurement. On the one hand, this would add a relevant contribution to better constrain the scalar interaction. On the other hand, in the framework of the SM, this result would improve by a factor of 2 the precision on Vud determined from the study of mirror transitions.

Summary and perspectives

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

The status of the first precision experiments in nuclear β decay performed with LPCTrap at GANIL has been described. Such studies are based on the use of a Paul trap to confine the radioactive ions and on the detection in coincidence of the β particles and the recoil ions. Significant results, especially the most precise value of aβν ever determined for a pure GT transition using a coincidence method, have been obtained. The last data are very promising and they should deliver the most precise absolute results on both the tensor and scalar sectors. A significant improvement of Vud extracted from the mirror transitions also seems reachable in an alternative interpretation of the 35Ar data. Moreover, an original recoil ion spectrometer enabled measurement of, for the first time, the charge-state distributions of recoiling ions produced by the decay of 1+ ions. The use of the sudden approximation in the theoretical approach has definitively been validated with the result obtained with the ‘ideal’ 6He1+ system, allowing unambiguous tests of the role of other processes in systems involving many electrons as in 35Ar1+ decay.

A new project involving 19Ne is now underway. This nucleus decays mainly through a mirror transition to the ground state of 19F (BR = 99.9858(20)%). The basic parameters (T1/2, BR, masses) involved in the decay being well known [9, 32], a value of aβν can be determined in the framework of the SM with sufficient precision. The precise measurement of aβν in this transition can again be interpreted either to constrain the exotic currents, CS and CT, in a global way, or to determine Vud. This experiment is a new challenge for LPCTrap. The half-life of 19Ne is indeed ten times larger than in the case of 35Ar and the recoil maximum kinetic energy is only 203 eV. These are extreme conditions that will push the limits of the device to new levels of sensitivity.

Very exciting opportunities will emerge in the coming years at GANIL with the development of new radioactive beams in an upgrade of SPIRAL [33] and in the forthcoming DESIR4 hall at SPIRAL2 [34]. The intensities reached during the first production tests for nuclei such as 23Mg, 25Al, 33Cl and 37K, which decay through mirror transitions, are encouraging and these beams should be available at GANIL in 2015.

  1. 1

    GANIL: Grand Accélérateur National d'Ions Lourds

  2. 2

    LIRAT: Ligne d'Ions RAdioactifs à Très basse énergie

  3. 3

    SPIRAL: Système de Production d'Ions Radioactifs Accélérés en Ligne

  4. 4

    DESIR: Décroissance Excitation et Stockage d'Ions Radioactifs

References

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies

Biographies

  1. Top of page
  2. Abstract
  3. The β−ν angular correlation
  4. The LPCTrap setup
  5. Measurements in the decay of 6He
  6. Measurements in the decay of 35Ar
  7. Summary and perspectives
  8. References
  9. Biographies
  • Image of creator

    Gilles Ban is presently Professor at Ecole Nationale Supérieure d'Ingénieurs de Caen and Laboratoire de Physique Corpusculaire de Caen. He got his PhD at the University of Paris in atomic physics and ion sources, he then worked with synchrotron light and X-ray optics. Since joining LPC Caen he has contributed to precision experiments at low energy such as correlations measurements or electric dipole moment searches with neutrons. Other interests are nuclear energy and medical applications.

  • Image of creator

    Dominique Durand is researcher at CNRS-IN2P3, and currently head of the Laboratoire de Physique Corpusculaire at Caen, France. He has obtained the Thibaud prize award (1997). His research interests have focused on nuclear collisions at intermediate energies (INDRA collaboration), double-beta decay experiments (NEMO-3 and SuperNEMO collaborations) and tests of the weak interaction through precision measurements at low energies.

  • Image of creator

    Xavier Fléchard is currently working at the LPC CAEN, France, as a CNRS researcher in the AMO division. After a PhD in atomic Physics at the University of Caen in the field of ion-atom collisions, he first worked at Kansas State University as a research associate on the development of the MOTRIMS technique to provide optically cooled and trapped atomic targets. His main research activities focus today on the use of atom and ion traps in specific fields of atomic and nuclear physics. They are shared between the study of ion-atom/molecules collisions, and precision measurements in nuclear beta decay to search for exotic interactions.

  • Image of creator

    Etienne Liénard is presently Associate Professor at the University of Caen and LPC Caen, France. He got his PhD at the Catholic University of Louvain, Belgium, in the field of nuclear reactions, using the first radioactive beam post-accelerated in the world. He held a first position as a research associate at GANIL, France, where he focused on neutron detection and halo nuclei studies. At LPC Caen, his main research activities concern precision measurements in nuclear beta decay to search for new physics beyond the Standard Model, and he is presently responsible of the LPCTrap setup installed at GANIL

  • Image of creator

    Oscar Naviliat-Cuncic is currently a faculty member at NSCL, Michigan State University. He held previously positions as Professor at the University of Caen, France, and as research associate at the Swiss Federal Institute of Technology in Zurich. After earning a PhD in nuclear spectroscopy at the Catholic University of Louvain, Belgium, his research interests have focused on tests of discrete symmetries and searches for new interactions through precision measurements at low energies. He has been involved in experiments measuring correlations using muons, neutrons and nuclei, some of which required the use traps for the confinement of particles.