Spectroscopy along Flerovium Decay Chains: Discovery of ^{280}Ds and an Excited State in ^{282}Cn.

A nuclear spectroscopy experiment was conducted to study α-decay chains stemming from isotopes of flerovium (element Z=114). An upgraded TASISpec decay station was placed behind the gas-filled separator TASCA at the GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt, Germany. The fusion-evaporation reactions ^{48}Ca+^{242}Pu and ^{48}Ca+^{244}Pu provided a total of 32 flerovium-candidate decay chains, of which two and eleven were firmly assigned to ^{286}Fl and ^{288}Fl, respectively. A prompt coincidence between a 9.60(1)-MeV α particle event and a 0.36(1)-MeV conversion electron marked the first observation of an excited state in an even-even isotope of the heaviest man-made elements, namely ^{282}Cn. Spectroscopy of ^{288}Fl decay chains fixed Q_{α}=10.06(1)  MeV. In one case, a Q_{α}=9.46(1)-MeV decay from ^{284}Cn into ^{280}Ds was observed, with ^{280}Ds fissioning after only 518  μs. The impact of these findings, aggregated with existing data on decay chains of ^{286,288}Fl, on the size of an anticipated shell gap at proton number Z=114 is discussed in light of predictions from two beyond-mean-field calculations, which take into account triaxial deformation.

A nuclear spectroscopy experiment was conducted to study α-decay chains stemming from isotopes of flerovium (element Z ¼ 114). An upgraded TASISpec decay station was placed behind the gas-filled separator TASCA at the GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt, Germany. The fusion-evaporation reactions 48 Ca þ 242 Pu and 48 Ca þ 244 Pu provided a total of 32 flerovium-candidate decay chains, of which two and eleven were firmly assigned to 286 Fl and 288 Fl, respectively. A prompt coincidence between a 9.60(1)-MeV α particle event and a 0.36(1)-MeV conversion electron marked the first observation of an excited state in an even-even isotope of the heaviest man-made elements, namely 282 Cn. Spectroscopy of 288 Fl decay chains fixed Q α ¼ 10.06ð1Þ MeV. In one case, a Q α ¼ 9.46ð1Þ-MeV decay from 284 Cn into 280 Ds was observed, with 280 Ds fissioning after only 518 μs. The impact of these findings, aggregated with existing data on decay chains of 286;288 Fl, on the size of an anticipated shell gap at proton number Z ¼ 114 is discussed in light of predictions from two beyond-mean-field calculations, which take into account triaxial deformation. DOI: 10.1103/PhysRevLett.126.032503 Experimental studies of superheavy atomic nuclei at the present limits of proton number Z and mass number A are of fundamental significance for chemistry and physics. Following the approval and naming of elements up to Z ¼ 118 (oganesson, Og) [1], the hunt for new elements with proton numbers Z ≥ 119 has intensified. In parallel, there are continued efforts to collect increasingly detailed information on chemical and physical properties of these rare species. Comprehensive overviews of the field, summarizing state-of-art experimental methods and theoretical approaches, can be found in, for instance, Refs. [2,3].
Elements 114 (flerovium, Fl) and 115 (moscovium, Mc) play a special role in the context of the detailed investigations. The main reason is that production cross sections for a number of their isotopes were firmly established to be rather high at σ prod ≈ 5-10 pb [4,5]. In the case of Fl, considerable efforts were undertaken to establish chemical properties [6][7][8][9]. In the case of Mc, primarily the decay chain associated with 288 Mc was subject to nuclear spectroscopy studies [10,11]. Supported by Monte Carlo simulations [12], decay schemes based on α-photon coincidences were proposed along the 288 Mc decay chain [10,13], followed by contemporary nuclear theory assessments [14]. More recently, the mass number A ¼ 288 was directly measured for that decay chain [15].
Nuclear spectroscopy results on Fl isotopes are relevant for nuclear theory. Most microscopic-macroscopic models see Z ¼ 114 as magic proton number, typically in combination with neutron number N ¼ 184, while a majority of modern mean-field approaches prefer Z ¼ 120 [2,16,17]. At spherical deformation, an extended region of low level density and shell corrections is expected for 114 ≤ Z ≤ 120 by most models. However, the heaviest flerovium isotope in experimental reach at present is 290 Fl with N ¼ 176, i.e., still eight neutrons short of the anticipated doubly magic core 298 Fl. Whether or not the isotopes in experimental reach are near spherical, well deformed, or show aspects of shape coexistence, is one of the key questions that remains to be tackled.
In this Letter, we report on newly observed decay properties along 286 Fl and 288 Fl decay chains stemming from a high-resolution spectroscopy experiment. Results from the same experiment on the odd-A isotope 289 Fl will be published separately [18]. Notes on preparatory efforts and some calibration procedures can be found in Ref. [19], while Ref. [20] provides further details on data analysis and statistical assessments. The Supplemental Material to this Letter [21] summarizes key information on previous and present results on the decay chains of interest.
The Universal Linear Accelerator (UNILAC) at the GSI Helmholtzzentrum für Schwerionenforschung, Darmstadt, Germany, delivered a beam of 48 Ca 10þ ions with a typical intensity of 5 × 10 12 particles per second, time averaged over its pulsed structure, 5 ms beam on and 15 ms beam off. A beam integral of 6.0ð4Þ × 10 18 ions was collected in two experimental runs. With an initial energy of 6.021ð2Þ MeV=u, then degraded by Ti foils with thicknesses indicated in Table I, the 48 Ca beam impinged on a rotating target wheel [22]. The latter comprised four segments of either enriched 242 Pu (> 99.5%) or 244 Pu (> 98.5%). The Pu material was deposited by molecular plating on 2.2-and 2.3-μm thick Ti foils [23]. The Ti foils faced the degraded beam. In the first run, a mixed target wheel with one 242 PuO 2 and three 244 PuO 2 segments was used. The second run had only 244 PuO 2 segments mounted onto the wheel. Table I provides an overview of thicknesses of degrader foils, backing foils, and target layers, as well as derived midtarget beam energies and beam integrals.
Behind the target wheel, the recoil separator TASCA [26] was used to direct reaction products toward an upgraded version [19] of the TASISpec decay station [27]. TASCA was filled with 0.8 mbar He gas and was set to center ions in the focal plane with a magnetic rigidity of Bρ ¼ 2.27 Tm [8,9,28]. Transmission of Fl fusion-evaporation residues from the target wheel into TASISpec was estimated to be 30(3)% [29,30].
The core of TASISpec is a cubelike arrangement of a 0.31-mm thick, 32 × 32-strip double-sided silicon strip detector (DSSD) and four additional 0.97 mm thick, 16 × 16-strip DSSDs placed upstream. A second 0.31-mm DSSD was placed behind the central implantation DSSD to veto background radiation from, for instance, β decays of transfer reaction products during beam-off periods [19]. All six DSSD wafers are 60 × 60 mm 2 in area, including a surrounding 1-mm wide guard ring. 80-μs traces of the 256 preamplified signals were recorded by 50-MHz, 14-bit sampling digitizers. Five composite germanium detectors were placed closely behind each of the five sides of the DSSD cube-one seven-crystal Cluster detector [31] downstream, and four novel four-crystal Compex detectors [32] behind each of the upstream DSSDs. Their signals were processed by 100-MHz, 16-bit sampling digitizers, while only flattop energy, baseline, time, and a pileup recognition flag were recorded.
List mode data were generated once a strip of the implantation detector registered a signal above a threshold of ≈150 keV, or alternatively when one of the four upstream DSSDs registered a signal in excess of ≈5 MeV. Trigger rates were typically ≈1200 (≈150) events per second during beam-on (beam-off) periods. In case the detection of a 9.5-10.3-MeV α particle in the DSSD cube engaged the same pixel of the implantation detector as a preceding recoil candidate within 20 s and during a beam-off period, an electrostatic chopper near the ion source was activated to bend the beam away from the axis within 20 μs. This allowed for low-background measurements of subsequent decays up to 200 or 300 s for the first or second part of the experiment, respectively. Event trigger type, beam on-off status, in-beam target segment, activated chopper, and various detector and event rates were recorded as well. Details on analysis aspects can be found in Refs. [19,20].
For the even-even isotopes 286;288 Fl, a search for timeand position-correlated recoil-αð−αÞ-fission sequences was Here, α-particle energies concern both full-energy α events in the implantation detector as well as those reconstructed between implantation and any upstream DSSD. Throughout the whole experiment, as few as 44 beam-off fission events were registered. Hence, these 44 events as well as candidate chains arising from the search were investigated in more detail, in particular for additional α-escape events between recoil and concluding fission, as well as photons or (conversion) electrons in prompt coincidence with any decay event. Taking into account the pulsed beam structure of UNILAC, the option for extended beam-off periods (see above), and the possibility of missing events due to data-acquisition dead time, a decision tree yielded a probability of ≈93% to identify
PHYSICAL REVIEW LETTERS 126, 032503 (2021) 032503-3 Material [21], which provide the spectral and statistical analyses of five data ensembles of the decays of 286;288 Fl and 282;284 Cn: (i) previous direct production; (ii) previous indirect production; (iii) combination of ensembles (i) and (ii); (iv) present data; and (v) combination of ensembles (iii) and (iv). Based on the results of the extended Schmidt test [43] shown in Table II in the Supplemental Material, all five ensembles and all assigned chains and decay steps within the respective ensemble were found to be consistent. One note is, however, in order: irrespective of either direct or indirect production of 288 Fl, there seemed to be hints for two closely lying decay paths (ΔE ≈ 0.1 MeV) of similar strength [cf. Figs. 2(a), 2(d), 2(g), Ref. [21] ] in all earlier data sets [45]. The present data set resolves that puzzling observation. Superior spectroscopic quality in combination with the comparatively large amount of data points yielded a single α-decay line at 9.92(1) MeV with a full-width at half maximum of 35 keV [cf. Fig. 2(j) Ref. [21] ], thus fixing the decay characteristics of 288 Fl [cf. Fig. 1(a) and Table II].
Two decay chains, number 08 assigned to 288 Fl and number 02 assigned to 286 Fl, are of particular interest. Their observed implantation and decay series are illustrated in Figs. 1(b) and 1(c).
Chain 08 is built upon a recoil implantation of 13.6 MeV, followed by a presumed full-energy α-decay event, E α1 ¼ 9.91ð1Þ MeV, after 2.408 s during a beam-off period. This event triggered the beam shut-off routine. After 21.3 ms, another presumed full-energy α-decay event. E α2 ¼ 9.33ð1Þ MeV, was detected, with a fission event following only 518 μs later. For this event, one fission fragment left the implantation detector (141 MeV signal) and reached an upstream detector (27 MeV signal). With the first α-decay energy and both first correlation times being consistent with 288 Fl and 284 Cn averages, we conclude to have for the first time observed an α-decay branch Ds, which is found to fission with T 1=2 ¼ 0.36ð 172 16 Þ ms [46]. The random probability of observing this recoil-α-αfission decay sequence in our experiment was as low as 3 × 10 −12 [20]. Together with existing information on 292 Lv [40][41][42], this decay chain firmly establishes the first highresolution Q α sequence across Z ¼ 114 (cf . Table II) [47]. Fl-284 Cn-280 Ds Q α series with a selection of predictions, namely from macroscopic-microscopic (MM) approaches [48,49], various nonrelativistic density-functional theories (DFT) with different interactions [50,51], a relativistic one [52], and a first fully triaxial beyond-mean-field (TBMF) calculation [53,54].
Notably, the measured Q α values pass smoothly through Z ¼ 114, which is at variance with some MM descriptions, for example [48], with a pronounced Z ¼ 114 shell gap and thus near-spherical shapes around N ¼ 174. Similarly, several axially symmetric DFT with the Skyrme (SLY4) and Gogny (D1S) interactions predict a "kink" of Q α at 288 Fl as well, although they do not predict a Z ¼ 114 gap. Since triaxial shapes are important in these nuclei [54,56], we extended the axially symmetric Gogny calculations of Ref. [51] to include triaxial shapes. The results, not displayed here, show the persistence of the kink indicating that triaxiality is at least not the only reason for the discrepancy with experimental data. In turn, MM predictions based on a Woods-Saxon potential [49], the unified energy-density Skyrme functional (UNEDF1) [50], and the triaxial relativistic mean-field (RMF) model [52], respectively, are schematically consistent with the smooth experimental trend across Z ¼ 114. However, slopes and absolute values remain different, as can be seen in Fig. 2(a).
The TBMF calculations [53] were performed with the Gogny force and provide a good description of the experimental data: the Q α values for 288 Fl and 292 Lv agree with the experiment, while the one for 284 Cn is somewhat too small. These calculations were performed using a configuration space of thirteen major harmonic oscillator  2. (a) Comparison of observed and predicted Q α values of the α-decay sequence 292 Lv, 288 Fl, and 284 Cn. The theoretical numbers stem from macroscopic-microscopic (MM) models [48,49], axially symmetric (SLY4 and UNEDF1 [50], Gogny [51]) and triaxial (RMF [52]) density functional theory (DFT), and a triaxial beyond-mean-field (TBMF) description [54]. (b) Proposed excited states in 282 Cn compared with predictions of two TBMF calculations [54,55]  shells. We note, however, that although this size is sufficient to reach relative convergence, it is not evident that also absolute convergence of the total binding energies, in particular for the more deformed lighter nuclei, was reached. All models showing an almost linear trend in Q α values also predict weakly deformed, near-prolate shapes with β ≈ 0.15. The absence of a kink indicates that no pronounced shell effect from crossing Z ¼ 114 is observed (at N ¼ 174). This may provide an upper limit of the size of the proposed Z ¼ 114 gap at this neutron number.
Hindrance factors (HF) along this decay sequence are listed in Table II, calculated with respect to a modified Geiger-Nuttall relation [44]. The HF values are close to unity, with 288 Fl mildly deviating. This indicates in general similar wave functions of the involved mother and daughter 0 þ 1 ground states. The detection of chain 02, shown in Fig. 1(c), started with a recoil implantation of 13.6 MeV, followed by a presumed α-decay event of 9.60(1) MeV after 52.6 ms during a beam-off period. This event triggered the beam shut-off routine. After 1.48 ms, a fission event concluded the chain, with 239 MeV measured in the implantation DSSD. The random probability of observing such a sequence in our experiment was 8 × 10 −7 [20]. While the recoil-α and α-SF correlation times are perfectly compatible with the half lives known for the decay chain of 286 Fl, the 9.60(1) MeV is ≈0.6 MeV below the present best value E α ð 286 FlÞ ¼ 10.19ð3Þ MeV [21].
The unique point of this decay chain is the detection of a 0.36(1) MeV signal in one pixel of the upstream detectors in prompt coincidence, Δt < 50 ns, with the 9.60(1)-MeV event in the implantation detector. Recoil implantation depth and pixel-to-pixel direction exclude the possibility of a to-be-reconstructed α-particle event [20]. Thus, the only explanation for the observed event is an α-(conversion)electron coincidence. Second, a study of all possible α-e − events provided a very conservative upper limit of < 1% probability for the observed 9.60(1)-0.36(1)-MeV coincidence being random [20]. Finally, ½9.60ð1Þ þ 0.36ð1Þ MeV þ B e;K;L ðCnÞ ≈ E α ð 286 FlÞ.
In conjunction with the very short correlation times, an explanation for this observation is an E α ≈ 9.6 MeV branch with b α ¼ 1=29 ≈ 3%, which connects the 0 þ ground state of 286 Fl with an excited state of 282 Cn, located at ≈0.6 MeV excitation energy (cf. Table II for Q α ). In this scenario, the 0.36(1) MeV e − is arising from a converted electromagnetic transition.
The hindrance factor [44] of a 0 þ 1 → 0 þ 2 branch turns out to be close to unity, while the observation of a finite Δl α ≠ 0 α-particle branch with ≈0.6 MeV less energy is considerably less likely. Therefore, spin-parity I π ¼ 0 þ is suggested for the excited state at ≈0.6 MeV. The excitation energy of the first 2 þ state in 282 Cn is expected at lower energy, in fact Eð2 þ 1 Þ ≪ 0.6 MeV [see, e.g., Fig. 2(b) and Refs. [48][49][50][51][52] ]. The presence of a 2 þ 1 state below the excited 0 þ 2 state implies that the decay of the latter proceeds with high probability via an E2 cascade passing through the 2 þ state. This is illustrated in Fig. 1(d), which presents this tentative decay scenario [20]. Conversion coefficients are calculated at α tot ≈ 2.6 and 0.33 for 220-and 400-keV [57], respectively [57]. Even in case both conversion electrons, most likely of L2 character, are emitted backward, atomic relaxation processes are likely to contribute some energy to the measured 9.60(1) MeV event in the implantation DSSD. The summed electron energy contribution is estimated to 0.03(2) MeV [58,59]. This leaves E α ¼ 9.57ð3Þ MeV [corresponding to Q α ¼ 9.71ð3Þ MeV] and yields Eð0 þ 2 Þ ¼ 0.62ð4Þ MeV and Eð2 þ 1 Þ ¼ 0.22ð4Þ MeV for a consistent description of this event. Figure 2(b) shows the observed and proposed excited states in 282 Cn at 0.62(4) and 0.22(4) MeV in comparison with the sequences of excited 2 þ 1 and 0 þ 2 states predicted by two contemporary triaxial beyond-mean-field approaches: Gogny TBMF is based on the effective finite-range densitydependent Gogny force and UNEDF1 TBMF relies on mapping the Skyrme functional to a separable Hamiltonian. Both sets of calculations restore the particle number and angular momentum symmetries and incorporate the mixing of triaxial shapes by means of the generator coordinate method (GCM). For details we refer to Refs. [53][54][55]. In fact, a fully triaxial approach is called for, since restricted axially symmetric calculations for superheavy nuclei predict oblate and prolate shapes at similar values of β. The question and quest for shape coexistence comes to mind, eventually by means of low-lying 0 þ states as observed, for instance, in 186 Pb [60] or 188;192 Po [61].
Both calculations predict smoothly increasing 2 þ 1 energies, i.e., decreasing quadrupole deformation with increasing neutron number. However, they differ regarding the location of the excited 0 þ 2 states. While calculated 0 þ 2 energies are continuously decreasing in the UNEDF1 TBMF approach, the Gogny TBMF calculations predict rather sudden changes of the 0 þ 2 energy from isotope to isotope, caused by the occurrence of several axially symmetric and triaxial configurations at low energy in these nuclei, similar to what is discussed for the chain of flerovium isotopes in Ref. [54]. Interestingly, just for 282 Cn, a rather low-lying 0 þ 2 state close to the experimental value is predicted [cf. Fig. 2(b)].
To conclude, an experiment was conducted behind the TASCA separator at GSI Darmstadt, aiming at highresolution decay spectroscopy of Z ¼ 114 flerovium isotopes. For even-even 288 Fl, a precise Q α value was derived. For the first time, an α-decay branch was observed for 284 Cn. This led to the discovery of 280 Ds and the first determination of a Q α series across Fl. This Q α series, now completed with a firm data point at 284 Cn, was found to provide a stringent test of model predictions and, in turn, on PHYSICAL REVIEW LETTERS 126, 032503 (2021) 032503-5 the strength of an anticipated (magic) shell gap at proton number Z ¼ 114. An excited state was observed in the even-even isotope 282 Cn, illustrating the experimental reach of detailed spectroscopy of the heaviest elements. The existence of the state provides yet another anchor point for nuclear theory, because it seems to require an understanding of both shape coexistence and shape transitions for the heaviest elements.
The results presented here are based on the experiment U310, which was performed at the beam line X8/TASCA at the GSI Helmholtzzentrum für Schwerionenforschung, Darmstadt (Germany) in the frame of FAIR Phase-0. The authors would like to thank the ion-source and accelerator staff at GSI for their supreme efforts. This work has received funding from the European Unions