betaFeedback

WorksPhysics and Astronomy

Precise Q value measurements of 112,113^{112,113}Ag and 115^{115}Cd with the Canadian Penning trap for evaluation of potential ultra-low Q value ββ-decays

Abstract

An ultra-low Q value ββ-decay can occur from a parent nuclide to an excited state in the daughter with QULQ_{UL} <1 keV. These decays are of interest for nuclear ββ-decay theory and as potential candidates in neutrino mass determination experiments. To date, only one ultra-low Q value ββ-decay has been observed – that of 115^{115}In with QβQ_β = 147(10) eV. A number of other potential candidates exist, but improved mass measurements are necessary to determine if the decays are energetically allowed and, in fact, ultra-low. We performed precise ββ-decay Q value measurements of 112,113^{112,113}Ag and 115^{115}Cd and combined them with nuclear energy level data for the daughter isotopes to determine if the potential UL Q value ββ-decay branches of 112,113^{112,113}Ag and 115^{115}Cd are energetically allowed and <1 keV. The Canadian Penning Trap at ANL was used to measure the cyclotron frequency ratios of singly-charged 112,113^{112,113}Ag and 115^{115}Cd ions with respect to their daughters. From these measurements, the ground-state ββ-decay Q values were obtained. The 112^{112}Ag, 113^{113}Ag, and 115^{115}Cd ββ-decay Q values were measured to be 3990.16(22) keV, 2085.7(4.6) keV, and 1451.36(34) keV, respectively. These results were compared to energies of excited states in 112^{112}Cd at 3997.75(14) keV, 113^{113}Cd at 2015.6(2.5) and 2080(10) keV, and 115^{115}In at 1448.787(9) keV, resulting in QULQ_{\textrm{UL}} values of –7.59(26) keV, 6(11) keV, and 2.57(34) keV, respectively. The potential UL Q value decays of 112^{112}Ag and 115^{115}Cd have been ruled out. 113^{113}Ag is still a possible candidate until a more precise measurement of the 2080(10) keV, 1/2+^{+} state of 113^{113}Cd is available. In the course of this work we have found the ground state mass of 113^{113}Ag reported in the 2020AME to be lower than our measurement by 69(17) keV (a 4σσ discrepancy).

Introduction

Nuclear β\beta-decays offer insight into the underlying weak interaction processes that govern them, and the in-medium effects that modify them due to their occurrence inside the atomic nucleus [1, 2]. The majority of unstable nuclides known to exist decay via allowed β\beta-decay and have relatively large Q values. This results in them having typically short lifetimes, and making them fairly straight-forward to observe. However, there are a number of isotopes with low β\beta-decay Q values and/or high forbiddenness, which results in them having much longer half-lives. These isotopes are important tools for applications such as direct neutrino mass determination experiments e.g. [4, 5, 6] [3], and radioactive dating e.g. [7]. It is also important to categorize and understand these rare decays since they can contribute to backgrounds in other rare event experiments, such as neutrinoless double β\beta-decay (0νββ0\nu\beta\beta) and dark matter searches [8]. They also provide a testing ground for nuclear theory under atypical conditions [9].

Ground-state to ground-state (gs–gs) β\beta-decays have Q values covering a wide energy range from ≈\approx2.5 keV up to ∼\sim10 MeV. However, under special circumstances, such as a decay from the ground state of the parent nuclide to a nearby excited state in the daughter, the Q value can be much lower. Such a decay, with QUL≲1Q_{\mathrm{UL}}\lesssim1 keV, is known as an ultra-low (UL) Q value β\beta-decay [10].

To date, the only known UL Q value β\beta-decay is that of the 115^{115}In(9/2+9/2^+) ground state to the 115^{115}Sn(3/2+3/2^+) first excited state. This decay was discovered by Cattadori, et al. at the Laboratori Nazionali del Gran Sasso in 2005 [11], and was later confirmed at the HADES underground laboratory in 2009 [12, 13]. In these experiments, the 115^{115}In(9/2+9/2^+) →\rightarrow 115^{115}Sn(3/2+3/2^+) β\beta-decay was inferred via the detection of the 497.5 keV γ\gamma-ray emitted from the t1/2=11t_{1/2}=11 ps, 3/2+3/2^+ state. Precise Penning trap measurements of the 115^{115}In – 115^{115}Sn mass difference by groups at Florida State University [14] and the University of Jyväskylä [15], combined with the precisely known energy of the 115^{115}Sn(3/2+3/2^+) state, showed that this decay is energetically allowed with a QULQ_{\mathrm{UL}} value of 147(10) eV1. Theoretical descriptions of this UL Q value decay showed significant discrepancies between the calculated and measured partial half-life [10, 18]. The identification of additional UL Q value decays, and measurements of their partial half-lives are required to aid further theoretical developments. Furthermore, UL Q value β\beta-decays have the potential to be new candidates for direct neutrino mass determination experiments, since the fraction of decays in the energy interval ΔE\Delta E close to the endpoint, which is relevant for the determination of the neutrino mass, goes as (ΔE/Q)3(\Delta E/Q)^3 [19].

Potential UL Q value decay branches were identified in 115Cd{}^{115}\mathrm{Cd} [20] and 135Cs{}^{135}\mathrm{Cs} [21] with QULQ_{\mathrm{UL}} values of −2.8(4.0)-2.8(4.0) keV and 0.5(1.1)0.5(1.1) keV respectively, limited by the uncertainties in the masses of the parent and/or daughter isotopes. Since Ref. [20] was published, new atomic mass data in the 2016 Atomic Mass Evaluation (AME2016) [16] gave an UL Q value for 115Cd{}^{115}\mathrm{Cd} of 3.1(0.7)3.1(0.7) keV4. Although this would indicate that this transition is not UL, we note that in AME2016 and AME2020 the 115Cd{}^{115}\mathrm{Cd} mass is determined entirely through a (d,p)(d,p) reaction measurement linking it to 114Cd{}^{114}\mathrm{Cd}. Since atomic masses obtained via nuclear reaction data are not always reliable, and because the mass of 114Cd{}^{114}\mathrm{Cd} was determined from an older mass spectrometry technique, a direct Penning trap measurement of the 115Cd{}^{115}\mathrm{Cd} Q value is called for. Recently, the ground-state Q value of 135Cs{}^{135}\mathrm{Cs} was measured with the JYFLTRAP Penning trap and the Q value of the potential UL decay branch was determined to be QUL=0.44(31)Q_{\mathrm{UL}} = 0.44(31) keV [23], showing that it is indeed energetically allowed with QUL<1Q_{\mathrm{UL}} < 1 keV.

Additional potential UL Q value decay candidates have been identified in the literature [24, 25, 26, 27]. Again, more precise mass data for the parent and/or daughter isotopes are required to determine whether these decays are energetically allowed and if their Q values are ultra-low. In collaboration with the LEBIT group at the National Superconducting Cyclotron Laboratory, we recently investigated and ruled out two candidates identified in Ref. [27]: 89Sr{}^{89}\mathrm{Sr} and 139Ba{}^{139}\mathrm{Ba} [28]. The JYFLTRAP group recently ruled out potential UL Q value decays in 72As{}^{72}\mathrm{As} [29], and showed that 159Dy{}^{159}\mathrm{Dy} does have an electron capture decay branch to an excited state in the daughter with a QULQ_{\mathrm{UL}} value of around 1 keV [30]. They are also investigating other candidates [31].

In this paper we present the first direct measurement of the 115Cd{}^{115}\mathrm{Cd} gs–gs Q value (QgsQ_{\mathrm{gs}}) and determination of the Q value of the potential UL decay branch identified in Ref. [20]. We also investigate two potential UL Q value candidates identified in Ref. [27]: 112,113Ag{}^{112,113}\mathrm{Ag}. Decay schemes for all three candidates are shown in Figure 1. For 112Ag{}^{112}\mathrm{Ag} and 115Cd{}^{115}\mathrm{Cd}, potential UL Q value decays are to the 1 or 2+2^+ level at 3997.8 keV in 112Cd{}^{112}\mathrm{Cd}, and to the 9/2+9/2^+ level at 1448.8 keV in 115In{}^{115}\mathrm{In}, respectively. In the case of 113Ag{}^{113}\mathrm{Ag}, two potential UL Q value branches to 1/2+1/2^+ states in 113Cd{}^{113}\mathrm{Cd} at 2015.6 keV and 2080 keV are shown. A decay to the lower energy state was identified as a potential UL Q value transition based on the gs–gs Q value obtained using mass data from AME2016 [16], but

Figure 1. (color online) Decay schemes for 112,113Ag{}^{112,113}\mathrm{Ag} and 115Cd{}^{115}\mathrm{Cd} showing the main β\beta-decay branches (solid black arrows) and the potential ultra-low Q value decay branches (dashed blue arrows) investigated in this work. The ground-state to ground-state Q values are obtained using data from the AME2020 [22]. All values are given in units of keV.

was ruled out by our measurement presented here. Based on our new result, a decay to the higher energy state was later identified as a potential UL Q value transition.

Experimental Description

The 112,113Ag{}^{112,113}\mathrm{Ag}, and 115Cd{}^{115}\mathrm{Cd} gs–gs Q values were obtained via measurements of the cyclotron frequency ratio of singly-charged ions of the parent and daughter isotopes using the Canadian Penning Trap (CPT) at Argonne National Laboratory (ANL) [32, 33].

CPT is currently coupled to the Californium Rare Isotope Breeder Upgrade (CARIBU) facility that produces neutron-rich rare isotopes via spontaneous fission of 252Cf{}^{252}\mathrm{Cf} [34]. Fission products are thermalized in a gas catcher [35] and extracted as singly- or doubly-charged ions. The ions are then accelerated and go through a high-resolution mass separator to select ions of a particular A/qA/q. The selected ions then enter into a radio-frequency quadrupole (RFQ) cooler and buncher to prepare low emittance bunches for injection into the CARIBU multi-reflection time-of-flight mass separator (MR-TOF-MS) [36] where the next stage of mass selection occurs as ions reflect between electrostatic mirrors and separate in time-of-flight with respect to their mass, providing typical mass resolving powers of up to 100,000. A Bradbury-Nielsen Gate (BNG) [37] is then used to select a particular isotope. However, depending on the mass difference between isobars, more than one species can pass through the BNG. Finally, ion bunches are accumulated and further cooled in a linear Paul trap before being injected into the Penning trap.

The CPT has a hyperbolic geometry with compensation ring and tube electrodes and is immersed in a uniform 5.9 T magnetic field produced by a superconducting solenoidal magnet. Ions confined in the Penning trap undergo three normal modes of motion: axial, reduced-cyclotron, and magnetron, with characteristic frequencies fzf_z, f+f_+ and f−f_-, respectively [38]. By combining measurements of the observable normal mode frequencies, one can obtain the free-space cyclotron frequency for an ion with charge-to-mass ratio q/mq/m in a magnetic field of strength BB:

fc=qB2πm.(1)f_c = \frac{qB}{2\pi m}. \tag*{(1)}

At the CPT, fcf_c is measured using the so-called Phase-Imaging Ion Cyclotron Resonance (PI-ICR) technique that was originally developed and implemented by the SHIP-TRAP group [39, 40]. This technique enables a measurement of the total phase accumulated by an ion in its reduced-cyclotron or magnetron motion during a precisely defined time interval, that in turn can be used to determine the ion’s frequency in the trap. The phase determination is performed by ejecting ions from the trap onto a position sensitive micro channel plate (MCP) detector, preserving the ion’s phase information.

In this work, the direct method for determining fcf_c described in Ref. [40] was used [33]. This method involves two separate phase measurements known as the reference spot measurement and the final spot measurement (an additional measurement at the start of the experiment is also required to determine the spot on the MCP that corresponds to the center of the trap). For both the reference and final spot measurements, the reduced-cyclotron motion of the ion is first excited to a well-defined radius via a pulsed rf dipole drive at a frequency close to f+f_+. For the reference spot, a quadrupole rf drive pulse at frequency frf≈f++f−=fcf_{\mathrm{rf}}\approx f_+ + f_- = f_c is immediately applied, which converts the ion’s reduced-cyclotron motion into magnetron motion. The ion is then allowed to accumulate phase in its magnetron motion for a specific time period before being ejected from the trap. For the final spot, the quadrupole rf drive pulse is applied after a phase accumulation period of length tacct_{\mathrm{acc}}, so that the ion accumulates mass-dependent reduced-cyclotron phase before its reduced-cyclotron motion is converted into magnetron motion. The ion then remains in its magnetron orbit for an additional period of time until it is ejected such that the total time spent in the trap during the reference and final spot measurements is the same. The cyclotron frequency, fcf_c, is determined from the total phase difference Δϕ\Delta\phi between the reference and measurement spots during the time interval tacct_{\mathrm{acc}} [40].

fc=Δϕ2πtacc=ϕmeas+2πN2πtacc,(2)f_c = \frac{\Delta\phi}{2\pi t_{\mathrm{acc}}} = \frac{\phi_{\mathrm{meas}} + 2\pi N}{2\pi t_{\mathrm{acc}}}, \tag*{(2)}

where ϕmeas\phi_{\mathrm{meas}} is the measured angle between the reference and final spot and NN is the number of complete revolutions for an ion with cyclotron frequency fcf_c during time tacct_{\mathrm{acc}}.

In order to obtain ϕmeas\phi_{\mathrm{meas}}, the central coordinates and associated uncertainties of the reference spot and final spot need to be determined. This was done using an unsupervised learning cluster-finding model, the Gaussian mixture model, which has been developed based on an expectation–maximization algorithm, see e.g. Ref. [41]. For the final phase measurement, there can be several spots along with the spot corresponding to the ion of interest depending on contaminants present in the beam. In such cases, tacct_{\mathrm{acc}} was carefully chosen so that the spot of interest is well separated from the other spots. Such an example can be seen in Figure 2, which displays the output of the cluster-finding model and shows the presence of A=113A=113 isobars 113Ag{}^{113}\mathrm{Ag}, 113mAg{}^{113m}\mathrm{Ag}, 113Cd{}^{113}\mathrm{Cd}, and 113Pd{}^{113}\mathrm{Pd}. Typically ∼100\sim100 – 500500 ions were accumulated in a spot to enable a determination of fcf_c to a precision of ∼3\sim3 – 55 mHz. Such a measurement took ∼2\sim2 – 3030 mins, depending on the rate at which the isotope of interest was delivered to the Penning trap. If necessary, this rate was limited to allow only a few ions per shot into the trap to avoid potential systematic frequency shifts due to ion-ion interactions.

Figure 2. (Color online) An example of a PI-ICR data set showing the separation of 113^{113}Ag (green spot) and the 113m^{113\mathrm{m}}Ag isomer (blue spot), as well as 113^{113}Cd and 113^{113}Pd (orange spot) isobaric contaminants, as identified with the cluster-finding algorithm. The red dot in each spot shows the center of each cluster. Grey dots represent counts on the MCP that are identified as noise. In this data set, a 550 ms tacct_{\mathrm{acc}} was used.

Data and Analysis

Experimental Runs and Data

The experimental data were acquired during three separate runs in 2016, 2017 and 2018. During the initial 2016 run, data were taken for the ratios of interest 112^{112}Ag+^{+}/ 112^{112}Cd+^{+}, 113^{113}Ag+^{+}/ 113^{113}Cd+^{+}, and 115^{115}Cd+^{+}/ 115^{115}In+^{+}, and for test ratios 112^{112}Sn+^{+}/ 112^{112}Cd+^{+} and 115^{115}In+^{+}/ 115^{115}Sn+^{+} that involve isotopes whose masses have already been precisely measured with Penning traps [14, 15, 42]. In this data we observed shifts in the QQ values calculated from the test ratio measurements of up to 10 keV compared to literature values. We also observed variations in the QQ values for different tacct_{\mathrm{acc}} times that were found to result from variations in fcf_c for different tacct_{\mathrm{acc}} times.

After the 2016 run, it was discovered that these shifts were due to a systematic sinusoidal variation of fcf_c as a function of tacct_{\mathrm{acc}} with a frequency corresponding to the magnetron frequency of ions in the trap. A corresponding systematic variation was observed in the final radial position of the ions on the MCP with a 90∘90^\circ phase shift compared to the fcf_c data, see Figure 3 and Ref. [33]. These observations indicate that ions injected into the trap had some initial magnetron motion with a reproducible amplitude and phase when they were initially confined in the trap, before their reduced-cyclotron motion was first driven by the pulsed rf dipole drive. This motion is then transferred to the final magnetron motion of the ions before they are ejected from the trap, with a phase that depends on the phase accumulated during the tacct_{\mathrm{acc}} period. Hence, the position of the final spot on the MCP is modified slightly resulting in a final radial position and phase, ϕmeas\phi_{\mathrm{meas}}, that depend on tacct_{\mathrm{acc}}. Following this discovery, a second experimental run was performed in 2017 with the A=115A = 115 ion pairs 115^{115}Cd+^{+}/ 115^{115}In+^{+} and 115^{115}In+^{+}/ 115^{115}Sn+^{+}, and a third experimental run was undertaken in 2018 to take additional A=115A = 115 data and data for the A=112A = 112 ion pairs 112^{112}Ag+^{+}/ 112^{112}Cd+^{+} and 112^{112}Sn+^{+}/ 112^{112}Cd+^{+}.

Figure 3. (Color online) Sinusoidal variation of fcf_c and radius of the projected orbit of the 115^{115}Cd ion spot observed on the MCP as a function of accumulation time, tacct_{\mathrm{acc}}, from the 2017 data set.

During the 2017 and 2018 runs, tacct_{\mathrm{acc}} was systematically varied to map out and account for the sinusoidal variation of fcf_c versus tacct_{\mathrm{acc}}. A sinusoidal fit to the data of the form

fc(t)=fc0+A0(t)sin⁡(2πf−t+ψ)(3)f_c(t) = f_{c0} + A_0(t)\sin(2\pi f_{-}t + \psi) \tag*{(3)}

was then performed. In this fit, A0(t)A_0(t) is the amplitude of the sine fit function, ψ\psi is a phase offset, and fc0f_{c0} is the baseline cyclotron frequency when no systematic shift occurs. Hence, fc0f_{c0} and its associated uncertainty are extracted. In this fit, f−f_{-} was constrained to the measured magnetron frequency. We note that, although the magnitude of the shift in ϕmeas\phi_{\mathrm{meas}} does not depend on tacct_{\mathrm{acc}}, A0A_0 does because fcf_c from Eqn. (2) goes as 1/tacc1/t_{\mathrm{acc}}.

The procedure to measure fc0f_{c0} in this way was then repeated with the other isotope for the parent-daughter pair so that the QQ value could be obtained, as discussed in section IV. Depending on the isotope pair, between one and three fc0f_{c0} measurements were performed for each isotope, alternating between the two.

After obtaining fc0f_{c0} for ions of the parent and daughter isotopes, the cyclotron frequency ratio, corresponding to the inverse mass ratio of the ions, was obtained:

R=fc0pfc0d=mdmp.(4)R = \frac{f_{c0}^{p}}{f_{c0}^{d}} = \frac{m_{d}}{m_{p}}. \tag*{(4)}

In the case that more than one fc0f_{c0} measurement was performed for parent and/or daughter isotope, neighboring fc0f_{c0} measurements for one ion were linearly interpolated to the time of the fc0f_{c0} measurement of the other ion to account for temporal magnetic field drifts. A weighted average of all resulting cyclotron frequency ratio measurements for a given ion pair was then obtained. The average cyclotron frequency ratios, after applying the systematic corrections discussed below, are given in Table I.

Systematic Corrections and Checks

During the experimental runs in 2017 and 2018, the data were obtained by measuring fcf_{c} as a function of tacct_{\mathrm{acc}} and extracting the baseline value, fc0f_{c0}, from a fit using Eqn. (3). This procedure enabled us to account for the phase dependent shift to fcf_{c} as a function of tacct_{\mathrm{acc}} so that it did not affect the cyclotron frequency ratio, RR. Hence there is no ΔRfin\Delta R_{\mathrm{fin}} correction to the 2017 or 2018 data in Table I.

For the 2016 data, measurements were performed for each ratio at specific tacct_{\mathrm{acc}} times, resulting in frequencies and corresponding ratios that did suffer from systematic shifts. In the case of 115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+}, we were able to use the parameters from the fit of Eqn. (3) to the 2017 A=115A = 115 data to correct the 2016 data. The correction, ΔRfin\Delta R_{\mathrm{fin}}, and the corrected ratio are shown in Table I. After this correction, there is good agreement between the uncorrected 2016 115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+} data and the data taken in 2017 and 2018. As such, we averaged 115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+} data from all three runs to obtain the average ratio listed in Table I to be used to determine the 115Cd^{115}\mathrm{Cd} gs-β\beta-decay Q value.

Correcting the 2016 A=115A = 115 data was possible because the 2017 data was taken within two months of the 2016 data and no significant changes to the CPT apparatus were made. Before the 2018 run, some 18 months after the 2017 run, a magnet quench had occurred, requiring the magnet to be re-energized, and a new voltage source for the Penning trap electrodes had been installed. As such, the parameters from the fit of Eqn. (3) to the 2018 data did not reproduce those from the 2017 data. Hence, we were not able to use the 2018 A=112A = 112 data to correct the 2016 112Ag+/112Cd+^{112}\mathrm{Ag}^{+}/^{112}\mathrm{Cd}^{+} data. We also did not take additional data at A=113A = 113 in 2018 because the 2016 data was sufficient to rule out the potential UL Q value decay branch. As such, we do not apply a correction, ΔRfin\Delta R_{\mathrm{fin}}, to the 2016 112,113Ag^{112,113}\mathrm{Ag} data. Instead, we include an uncertainty due to this effect that we estimate from Eqn. (2) and Eqn. (3) based on the tacct_{\mathrm{acc}} that was used and the variation in the observed orbital radius of the ion spot on the MCP, as seen, for example, in Figure 3. We also applied the ΔRfin\Delta R_{\mathrm{fin}} correction to the 2016 115In+/115Sn+^{115}\mathrm{In}^{+}/^{115}\mathrm{Sn}^{+} test ratio data, and included the systematic uncertainty due to this effect in the 112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/^{112}\mathrm{Cd}^{+} test ratio, as reported in Table II.

After the 2018 run, a smaller, additional systematic shift to fcf_{c} data taken using the PI-ICR technique at the CPT was discovered [33]. This shift affected the phase of the reference spot and is due to contaminant ions of the same nominal A/qA/q in the trap. During the pulsed rf dipole drive at f+f_{+} that is used to initially drive the ions to their reduced cyclotron orbit, and also during the pulsed rf quadrupole drive at fcf_{c} that is used to convert cyclotron motion into magnetron motion, ions accumulate a phase difference that depends on m/qm/q. Because the f+f_{+} and fcf_{c} drive pulses are of short duration (≈500\approx500 μs\mu\mathrm{s}), the resulting phase difference is typically small (∼5∘\sim5^\circ), and a separation of different species in the reference spot is not observed. However, depending on the proportion of contaminant ions to the ion of interest entering the trap, the weighted average phase of the reference spot can be systematically offset from the phase of just the ions of interest [33]. This effect can be corrected for by determining the percentage of contaminant ions vs ions of interest and calculating the corresponding weighted phase shift. This correction was accounted for in the data, and has been included in Table I as the correction ΔRref\Delta R_{\mathrm{ref}}.

Our cyclotron frequency ratio measurements of 112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/^{112}\mathrm{Cd}^{+} and 115In+/115Sn+^{115}\mathrm{In}^{+}/^{115}\mathrm{Sn}^{+} were performed to serve as an independent check of our measurements by comparing them to the inverse mass ratios calculated with data listed in the AME2020 [22]. We also observed the 113mAg^{113\mathrm{m}}\mathrm{Ag} isomer, so were able to obtain the ratio 113mAg+/113Ag+^{113\mathrm{m}}\mathrm{Ag}^{+}/{}^{113}\mathrm{Ag}^{+}, which can be compared to the mass ratio of the 113Ag^{113}\mathrm{Ag} ground and isomeric state. A comparison of these data is shown in Table II.

Table II. Comparison of test ratio measurements, Rˉ\bar{R}, given in Table I to inverse mass ratios, RlitR_{\mathrm{lit}}, obtained from literature values for mass [22] and isomeric energies [43]. ΔR\Delta R is the difference (Rˉ−Rlit)×109(\bar{R}-R_{\mathrm{lit}})\times10^{9} with total uncertainty in parentheses.

Ion PairYearRlitR_{\mathrm{lit}}ΔR\Delta R (×10−9\times10^{-9})
112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/{}^{112}\mathrm{Cd}^{+}20160.999 981 582 5(37)-137(49)
112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/{}^{112}\mathrm{Cd}^{+}2018-0.3(4.4)
113Ag+/113mAg+^{113}\mathrm{Ag}^{+}/{}^{113\mathrm{m}}\mathrm{Ag}^{+}20160.999 999 586 4(10)30(44)
113Ag+/113mAg+^{113}\mathrm{Ag}^{+}/{}^{113\mathrm{m}}\mathrm{Ag}^{+}20169(13)
115In+/115Sn+^{115}\mathrm{In}^{+}/{}^{115}\mathrm{Sn}^{+}20170.999 995 351 9(2)-0.2(4.4)
115In+/115Sn+^{115}\mathrm{In}^{+}/{}^{115}\mathrm{Sn}^{+}2018-19(7)

Table 2.

Results and Discussion

Qgs values for 112,113Ag and 115Cd -decay

The goal of this work was to obtain gs–gs β\beta-decay Q values for 112,113Ag^{112,113}\mathrm{Ag} and 115Cd^{115}\mathrm{Cd}. These Q values, defined as the energy equivalent of the mass difference between parent and daughter atoms, can be determined from the measured cyclotron frequency ratios via

Qgs=(Mp−Md)c2=(Md−me)(Rˉ−1−1)c2,(5)Q_{\mathrm{gs}}=(M_{\mathrm{p}}-M_{\mathrm{d}})c^{2}=(M_{\mathrm{d}}-m_{\mathrm{e}})(\bar{R}^{-1}-1)c^{2}, \tag*{(5)}

where MpM_{\mathrm{p}} and MdM_{\mathrm{d}} are the mass of the parent and daughter atoms, respectively, and mem_{\mathrm{e}} is the mass of the electron. The conversion factor from atomic mass units to keV, 1 u=931 494.102 42(28) keV/c21\ \mathrm{u}=931\,494.102\,42(28)\ \mathrm{keV}/c^{2} from Ref. [44] was used. The average cyclotron frequency ratios listed in Table I were used to obtain these Q values, and the results are listed in Table III. Daughter atomic masses were taken from the AME2020 [22], and mem_{\mathrm{e}} from the most recent CODATA recommended values of the fundamental physical constants [44]. Q values obtained in this work are compared to those from the AME2020 in Table III and in Figure 4.

Ground-state to ground-state Q values measured in this work

Figure 4. (Color online) Ground-state to ground-state Q values measured in this work. The red bands show the AME2020 uncertainty and the red dots are our measured values.

Our Q values for 112Ag^{112}\mathrm{Ag} and 115Cd^{115}\mathrm{Cd} are slightly lower, but in agreement with the values obtained from the AME2020. Our result for 112Ag^{112}\mathrm{Ag} is a factor of 10 more precise, while that for 115Cd^{115}\mathrm{Cd} is about a factor of two more precise. Our result for 113Ag^{113}\mathrm{Ag} shows a significant, 4σ4\sigma discrepancy compared to the AME2020, and is a factor of three more precise.

Evaluation of QUL values for 112,113Ag and 115Cd

Using the new Q values listed in Table 3, the potential UL Q values for decay branches to excited states in the daughter nuclei can be evaluated via

Table III. Q values obtained in this work, QCPTQ_{\mathrm{CPT}}, from the cyclotron frequency ratio measurements listed in Table I and using Eqn. (5). Results obtained from the AME2020 [22], QAMEQ_{\mathrm{AME}}, and the differences ΔQ=QCPT−QAME\Delta Q=Q_{\mathrm{CPT}}-Q_{\mathrm{AME}} are also listed. All values are in keV.

DecayQCPTQ_{\mathrm{CPT}}QAMEQ_{\mathrm{AME}}ΔQ\Delta Q
112Ag→112Cd^{112}\mathrm{Ag}\rightarrow{}^{112}\mathrm{Cd}3990.16(22)3991.1(2.4)-1.0(2.4)
113Ag→113Cd^{113}\mathrm{Ag}\rightarrow{}^{113}\mathrm{Cd}2085.7(4.6)2016.5(16.6)69.2(17.3)
115Cd→115In^{115}\mathrm{Cd}\rightarrow{}^{115}\mathrm{In}1451.36(34)1451.88(65)-0.52(73)

Table 3.

QUL=Qgs−E∗,(6)Q_{\mathrm{UL}}=Q_{\mathrm{gs}}-E^{*}, \tag*{(6)}

where QgsQ_{\mathrm{gs}} is the gs–gs Q value from Table III, and E∗E^{*} is the energy of the final state in the daughter, from Ref. [45]. The QULQ_{\mathrm{UL}} values that we obtained are listed in Table IV.

Our new QULQ_{\mathrm{UL}} values definitively show that the potential UL Q value decay for 112Ag^{112}\mathrm{Ag} identified in Ref. [27] is not energetically allowed, and that the potential UL Q value decay for 115Cd^{115}\mathrm{Cd} identified in Ref. [20] is not <1<1 keV and hence not ultra-low. The fact that our new QgsQ_{\mathrm{gs}} value for 113Ag^{113}\mathrm{Ag} deviates from the AME value by 69 keV, means that the potential UL Q value decay to the 1/2+1/2^{+} state in 113Cd^{113}\mathrm{Cd} at 2015.6 keV is ≫1\gg1 keV and is ruled out. However, there is another 1/2+1/2^{+} state in 113Cd^{113}\mathrm{Cd} at 2080(10) keV that an UL Q value decay could potentially go to. Currently, the 10 keV uncertainty on the energy of this state is too large to say anything more definitive.

DecayE∗E^{*}QULQ_{\mathrm{UL}} CPTQULQ_{\mathrm{UL}} AME
112Ag→112Cd^{112}\mathrm{Ag} \rightarrow{}^{112}\mathrm{Cd}3997.75(14)-7.59(26)-6.6(2.4)
113Ag→113Cd^{113}\mathrm{Ag} \rightarrow{}^{113}\mathrm{Cd}2015.6(2.5)70.1(5.2)0.9(16.8)
2080(10)5.7(11.0)-63.5(19.4)
115Cd→115In^{115}\mathrm{Cd} \rightarrow{}^{115}\mathrm{In}1448.787(9)2.57(34)3.1(0.7)

Table 4. Q values for potential UL decay branches identified in Refs. [20, 27]. The QULQ_{\mathrm{UL}} values were obtained using the gs–gs Q values, QCPTQ_{\mathrm{CPT}} and QAMEQ_{\mathrm{AME}}, from Table III, and excited state energies, E∗E^{*} from Ref. [45]. All values listed are in keV.

Mass Excesses for 112,113^{112,113}Ag and 115^{115}Cd

The ratios in Table 1 were used to obtain absolute atomic masses for the parent nuclides, 112,113^{112,113}Ag, and 115^{115}Cd, via

Table 1. Average cyclotron frequency ratios obtained from the measurements performed in this work. Measurements were performed in three separate experimental runs in 2016, ’17 and ’18. ΔRfin\Delta R_{\mathrm{fin}} and ΔRref\Delta R_{\mathrm{ref}} are corrections (and associated uncertainties in parentheses) ×109\times10^{9} applied to the ratio to account for systematic shifts to the final and reference spots, respectively (see text for details). σst\sigma_{\mathrm{st}} is the statistical uncertainty ×109\times10^{9}, and Rˉ\bar{R} is the resulting corrected ratio, with combined statistical and systematic uncertainties in parentheses.

Ion PairYr.ΔRfin\Delta R_{\mathrm{fin}}ΔRref\Delta R_{\mathrm{ref}}σst\sigma_{\mathrm{st}}Rˉ\bar{R}
112Ag+/112Cd+^{112}\mathrm{Ag}^{+}/^{112}\mathrm{Cd}^{+}’160(48)48(5)80.999 961 658(49)
112Ag+/112Cd+^{112}\mathrm{Ag}^{+}/^{112}\mathrm{Cd}^{+}’180(0)-1.6(2)2.10.999 961 721 5(21)
112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/^{112}\mathrm{Cd}^{+}’160(48)18(2)90.999 981 445(49)
112Sn+/112Cd+^{112}\mathrm{Sn}^{+}/^{112}\mathrm{Cd}^{+}’180(0)2.1(2)2.40.999 981 582 2(24)
113Ag+/113Cd+^{113}\mathrm{Ag}^{+}/^{113}\mathrm{Cd}^{+}’160(42)2.3(2)120.999 980 169(44)
113Ag+/113mAg+^{113}\mathrm{Ag}^{+}/^{113\mathrm{m}}\mathrm{Ag}^{+}’160(41)7.5(8)150.999 999 616(44)
115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+}’16-4(9)15(2)1.40.999 986 451 4(93)
115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+}’170(0)-19(2)6.00.999 986 435 5(63)
115Cd+/115In+^{115}\mathrm{Cd}^{+}/^{115}\mathrm{In}^{+}’180(0)-1.4(2)3.90.999 986 440 1(39)
Average0.999 986 440 3(31)
115In+/115Sn+^{115}\mathrm{In}^{+}/^{115}\mathrm{Sn}^{+}’1626(13)5(1)2.20.999 995 361(13)
115In+/115Sn+^{115}\mathrm{In}^{+}/^{115}\mathrm{Sn}^{+}’170(0)14(2)3.90.999 995 351 8(44)
115In+/115Sn+^{115}\mathrm{In}^{+}/^{115}\mathrm{Sn}^{+}’180(0)4.9(5)7.00.999 995 332 7(70)

Table 1.

Mp=(Md−me)Rˉ−1+me,(7)M_p = (M_d - m_e)\bar{R}^{-1} + m_e, \tag*{(7)}

with corresponding values for MdM_d taken from Ref. [22]. Mass excesses were then obtained and are listed in Table V where they are compared with the values from the AME2020 [22].

The mass excesses for 112^{112}Ag and 115^{115}Cd are in good agreement with the AME2020 data, but are factors of approximately seven and two more precise, respectively. Our result for 113^{113}Ag shows that it is less bound by 69.2(17.2) keV compared to the AME2020 value, a 4σ4\sigma discrepancy. The mass of 113^{113}Ag in AME2020 is derived almost entirely from three 113^{113}Ag β\beta-decay measurements [46, 47], the most precise of which is listed as a private communication to the Nuclear Data Group from 1957 [22]. The 115^{115}Cd mass, on the other hand is derived from a (d,p)(d,p) reaction linking it to 114^{114}Cd, and the 112^{112}Ag mass is from a Penning trap measurement performed by the ISOLTRAP group [48]. We note that we did observe the 113m^{113m}Ag isomer in addition to the 113^{113}Ag ground state, as shown in Figure 2, and determined their mass difference to be 40.4(4.6) keV/c2c^2, consistent with the literature value of 43.5(1) keV [43] for the energy of the isomeric state.

IsotopeThis work (keV/c2c^2)AME2020 (keV/c2c^2)Δ\DeltaME (keV/c2c^2)
112^{112}Ag−86-86 584.70(33)−86-86 583.7(2.4)-1.0(2.4)
113^{113}Ag−86-86 957.6(4.6)−87-87 026.8(16.6)69.2(17.2)
115^{115}Cd−88-88 085.00(42)−88-88 084.5(0.7)-0.5(0.8)

Table 5. Mass excesses for 112^{112}Ag, 113^{113}Ag and 115^{115}Cd obtained in this work along with results from the AME2020 [22] and the difference Δ\DeltaME = MECPT_{\mathrm{CPT}} – MEAME_{\mathrm{AME}}.

Conclusion

We have performed precise determinations of the gs–gs β\beta-decay Q values of 112,113^{112,113}Ag, and 115^{115}Cd by measuring the cyclotron frequency ratios of singly-charged parent and daughter ions with the Canadian Penning Trap mass spectrometer. By comparing these Q values with excited state energy levels in the daughter nuclei, the QULQ_{\mathrm{UL}} values for potential UL Q value decays of 112^{112}Ag and 115^{115}Cd were found to be –7.59(26) keV and 2.57(34) keV respectively. The former is not energetically allowed and the latter is too large (i.e. >1> 1 keV) to be considered as an UL decay, ruling both out as potential UL Q value β\beta-decays. Our Q value measurement of 113^{113}Ag indicated a 69(17) keV discrepancy compared to data from the AME2020. This result ruled out the potential UL Q value decay to the 1/2+1/2^{+}, 2015.6 keV state in 113^{113}Cd, but indicated a new potential UL Q value branch to the 1/2+1/2^{+} at 2080(10) keV. Hence, 113^{113}Ag is still a potential candidate with a QULQ_{\mathrm{UL}} of 6(11) keV, but a more precise determination of the 113^{113}Cd(1/2+1/2^{+}, 2080 keV) state energy will be required to further evaluate this decay branch.

We also report improved atomic masses for 112,113^{112,113}Ag and 115^{115}Cd. The atomic masses of 112^{112}Ag and 115^{115}Cd are in good agreement with the AME2020 values, and have had their uncertainties reduced by factors of seven and two, respectively. We observed an ≈70\approx70 keV discrepancy in the mass of 113^{113}Ag compared to the AME.

Acknowledgments

This material is based upon work supported by the US Department of Energy, Office of Science, Office of Nuclear Physics under contract No. DE-AC02-06CH11357 and Award No. DE-SC0015927, by NSERC (Canada), Application No. SAPPI-2015-00034 and SAPPI-2018-00028, and by the National Science Foundation under grant No. PHY-1713857. Support was also provided by Central Michigan University.

References

References

  1. [4]L. Gastaldo et al. (ECHo Collaboration), The European Physical Journal Special Topics 226, 1623 (2017).
  2. [5]B. Alpert et al. (HOLMES Collaboration), The European Physical Journal C 75, 112 (2015).
  3. [6]C. Arnaboldi, C. Brofferio, O. Cremonesi, E. Fiorini, C. Lo Bianco, L. Martensson, A. Nucciotti, M. Pavan, G. Pessina, S. Pirro, E. Previtali, M. Sisti, A. Giuliani, B. Margesin, and M. Zen, Phys. Rev. Lett. 91, 161802 (2003).
  4. [7]V. P. Chechev, Phy. At. Nucl. 74, 1713 (2011).
  5. [8]J. Kostensalo and J. Suhonen, Int. J. Mod. Phys. A 33, 1843008 (2018).
  6. [9]M. T. Mustonen and J. Suhonen, AIP Conference Proceedings 1304, 401 (2010).
  7. [10]M. T. Mustonen and J. Suhonen, J. Phys. G: Nucl. Part. Phys. 37, 064008 (2010).
  8. [11]C. M. Cattadori, M. D. Deo, M. Laubenstein, L. Pandola, and V. I. Tretyak, Nucl. Phys. A 748, 333 (2004).
  9. [12]J. S. E. Wieslander, J. Suhonen, T. Eronen, M. Hult, V.-V. Elomaa, A. Jokinen, G. Marissens, M. Misiasek, M. T. Mustonen, S. Rahaman, C. Weber, and J. Äystö, Phys. Rev. Lett. 103, 122501 (2009).
  10. [13]E. Andreotti, M. Hult, R. González de Orduña, G. Marissens, J. S. E. Wieslander, and M. Misiasek, Phys. Rev. C 84, 044605 (2011).
  11. [14]B. J. Mount, M. Redshaw, and E. G. Myers, Phys. Rev. Lett. 103, 122502 (2009).
  12. [15]J. S. E. Wieslander, J. Suhonen, T. Eronen, M. Hult, V.-V. Elomaa, A. Jokinen, G. Marissens, M. Misiasek, M. T. Mustonen, S. Rahaman, C. Weber, and J. Äystö, Phys. Rev. Lett. 103, 122501 (2009).
  13. [16]M. Wang, G. Audi, F. Kondev, W. Huang, S. Naimi, and X. Xu, Chin. Phys. C 41, 030003 (2017).
  14. [17]V. A. Zheltonozhsky, A. M. Savrasov, N. V. Strilchuk, and V. I. Tretyak, Europhys. Lett. 121, 12001 (2018).
  15. [18]J. Suhonen and M. Mustonen, Progress in Particle and Nuclear Physics 64, 235 (2010), neutrinos in Cosmology, in Astro, Particle and Nuclear Physics.
  16. [19]E. Ferri, D. Bagliani, M. Biasotti, G. Ceruti, G. D. Corsini, M. Faverzani, F. Gatti, A. Giacherо, C. Gotti, C. Kilbourne, et al., Phys. Procedia 61, 227 (2015).
  17. [20]M. Haaranen and J. Suhonen, The European Physical Journal A 49, 93 (2013).
  18. [21]M. Mustonen and J. Suhonen, Physics Letters B 703, 370 (2011).
  19. [22]M. Wang, W. Huang, F. Kondev, G. Audi, and S. Naimi, Chinese Physics C 45, 030003 (2021).
  20. [23]A. de Roubin, J. Kostensalo, T. Eronen, L. Canete, R. P. de Groote, A. Jokinen, A. Kankainen, D. A. Nesterenko, I. D. Moore, S. Rinta-Antila, J. Suhonen, and M. Vilén, Phys. Rev. Lett. 124, 222503 (2020).
  21. [24]M. T. Mustonen and J. Suhonen, AIP Conf. Proc. 1304, 401 (2010).
  22. [25]J. Suhonen, Physica Scripta 89, 054032 (2014).
  23. [26]J. Kopp and A. Merle, Phys. Rev. C 81, 045501 (2010).
  24. [27]N. D. Gamеz, R. Bhandari, M. H. Gamage, R. Sandler, and M. Redshaw, Hyp. Int. 240, 43 (2019).
  25. [28]R. Sandler, G. Bollen, N. D. Gamage, A. Hamaker, C. Izzo, D. Puentes, M. Redshaw, R. Ringle, and I. Yandow, Phys. Rev. C 100, 024309 (2019).
  26. [29]Z. Ge, T. Eronen, A. de Roubin, D. A. Nesterenko, M. Hukkanen, O. Beliuskina, R. de Groote, S. Geldhof, W. Gins, A. Kankainen, A. Koszorús, J. Kotila, J. Kostensalo, I. D. Moore, A. Raggio, S. Rinta-Antila, J. Suhonen, V. Virtanen, A. P. Weaver, A. Zadvor­naya, and A. Jokinen, Phys. Rev. C 103, 065502 (2021).
  27. [30]Z. Ge, T. Eronen, K. S. Tyrin, J. Kotila, J. Kostensalo, D. A. Nesterenko, O. Beliuskina, R. de Groote, A. de Roubin, S. Geldhof, W. Gins, M. Hukkanen, A. Jokinen, A. Kankainen, A. Koszorús, M. I. Krivoruchenko, S. Kujanpää, I. D. Moore, A. Raggio, S. Rinta-Antila, J. Suhonen, V. Virtanen, A. P. Weaver, and A. Zadvor­naya, Phys. Rev. Lett. 127, 273201 (2021).
  28. [31]A. Kankainen, T. Eronen, D. Nesterenko, A. de Roubin, and M. Vilén, Hyperfine Interactions 241, 43 (2020).
  29. [32]G. Savard, R. Barber, C. Boudreau, F. Buchinger, J. Caggiano, J. Clark, J. Crawford, H. Fukutani, S. Gulick, J. Hardy, et al., Hyperfine Interactions 132, 223 (2001).
  30. [33]R. Orford, J. Clark, G. Savard, A. Aprahamian, F. Buchinger, M. Burkey, D. Gorelov, J. Klimes, G. Morgan, A. Nystrom, et al., Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 463, 491 (2020).
  31. [34]G. Savard, S. Baker, C. Davids, A. Levand, E. Moore, R. Pardo, R. Vondrasek, B. Zabransky, and G. Zinkann, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 266, 4086 (2008).
  32. [35]G. Savard, J. Clark, C. Boudreau, F. Buchinger, J. Crawford, H. Geissel, J. Greene, S. Gulick, A. Heinz, J. Lee, A. Levand, M. Maier, G. Münzenberg, C. Scheidenberger, D. Seweryniak, K. Sharma, G. Sprouse, J. Vaz, J. Wang, B. Zabransky, and Z. Zhou, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 204, 582 (2003).
  33. [36]T. Y. Hirsh, N. Paul, M. Burkey, A. Aprahamian, F. Buchinger, S. Caldwell, J. A. Clark, A. F. Levand, L. L. Ying, S. T. Marley, G. E. Morgan, A. Nystrom, R. Orford, A. P. Galván, J. Rohrer, G. Savard, K. S. Sharma, and K. Siegl, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 376, 229 (2016), proceedings of the XVIth International Conference on Electromagnetic Isotope Separators and Related Topics (EMIS2015), Grand Rapids, MI, U.S.A., 11-15 May 2015.
  34. [37]N. E. Bradbury and R. A. Nielsen, Phys. Rev. 49, 388 (1936).
  35. [38]L. S. Brown and G. Gabrielse, Rev. Mod. Phys. 58, 233 (1986).
  36. [39]S. Eliseev, K. Blaum, M. Block, C. Droese, M. Goncharov, E. Minaya Ramirez, D. Nesterenko, Y. N. Novikov, and L. Schweikhard, Physical Review Letters 110, 082501 (2013).
  37. [40]S. Eliseev, K. Blaum, M. Block, A. Dörr, C. Droese, T. Eronen, M. Goncharov, M. Höcker, J. Ketter, E. M. Ramirez, et al., Applied Physics B 114, 107 (2014).
  38. [41]C. Weber, D. Ray, A. Valverde, J. Clark, and K. Sharma, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 1027, 166299 (2022).
  39. [42]S. Rahaman, V.-V. Elomaa, T. Eronen, J. Hakala, A. Jokinen, A. Kankainen, J. Rissanen, J. Suhonen, C. Weber, and J. Äystö, Phys. Rev. Lett. 103, 042501 (2009).
  40. [43]F. Kondev, M. Wang, W. Huang, S. Naimi, and G. Audi, Chinese Physics C 45, 030001 (2021).
  41. [44]E. Tiesinga, P. J. Mohr, D. B. Newell, and B. N. Taylor, Rev. Mod. Phys. 93, 025010 (2021).
  42. [45]“National nuclear data center,” https://www.nndc.bnl.gov/.
  43. [46]Z.-I. Matumoto and T. Tamura, Journal of the Physical Society of Japan 29, 1116 (1970).
  44. [47]B. Fogelberg, Y. Zongyuan, B. Ekström, E. Lund, K. Aleklett, and L. Sihver, Zeitschrift für Physik A Atomic Nuclei 337, 251 (1990).
  45. [48]M. Breitenfeldt, C. Borgmann, G. Audi, S. Baruah, D. Beck, K. Blaum, C. Böhm, R. B. Cakirli, R. F. Casten, P. Delahaye, M. Dorschwak, S. George, F. Herfurth, A. Herlert, A. Kellerbauer, M. Kowalska, D. Lunney, E. Minaya-Ramirez, S. Naimi, D. Neidherr, M. Rosenbusch, R. Savreux, S. Schwarz, L. Schweikhard, and C. Yazidjian, Phys. Rev. C 81, 034313 (2010).

Paper details

vlnc:W_6g9j33cx
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents