Recent progress in experiments on the heaviest nuclides at SHIP

Understanding the atomic and nuclear properties of very heavy and superheavy nuclides is one of the main objectives in experiments at the recoil separator SHIP at the GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt, Germany. This article summarizes recent experiments performed on isotopes of the elements with Z = 102–105. The radionuclides of interest were produced in heavy-ion fusion-evaporation reactions with lead and bismuth targets and separated from the primary beam by the velocity filter SHIP. The studies by α\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha $$\end{document}–γ\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\gamma $$\end{document} spectroscopy experiments, direct mass measurements and laser spectroscopy employed the SHIP decay-spectroscopy setup, the new COMPASS detector, the mass spectrometer SHIPTRAP, and the laser spectroscopy setup RADRIS. This article briefly introduces the experimental approaches and discusses selected results providing a more comprehensive picture of the nuclear structure around the deformed neutron shell closure at N=152\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N=152$$\end{document}. Future perspectives to extend such experiments towards other regions in the heaviest nuclei will also be addressed.


Introduction
The GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt, Germany, is a center for accelerator-based fundamental research which was established more than 50 years ago. One of the driving scientific motivations at that time was the search for superheavy elements (SHE). The existence of these nuclei with Z > 103 was predicted in the late 1960s [1,2] based on theoretical models extending the nuclear shell model [3,4]. These models explained the stability of superheavy nuclei (SHN) by nuclear shell effects that provide an increased binding energy by as much as 10 MeV to stabilize them against spontaneous fission. One of the most striking features in the early predictions was the existence of a region of superheavy nuclei near spherical proton and neutron shell closures at Z = 114 and N = 184 [1,2] where the half-lives would exceed 10 9 years (with a partial fission half-life of 10 16 years) [5]. This region was coined island of stability.
Reaching this island motivated significant experimental efforts in the following decades that resulted in the discovery of new elements up to oganesson with Z = 118 [6][7][8]. Evidence confirming the existence of the island of stability was indeed obtained from increasing half-lives of more neutron-rich superheavy isotopes [8]. However, more recent theoretical predictions suggest that the longest half-lives on this island will only be on the order of 100 years [9]. This is still remarkable for very exotic nuclides, but it rules out that SHE exist in nature. Nonetheless, SHE remain a forefront topic of nuclear physics research and many scientific questions remain open. Some of these questions and the theoretical description of SHE are discussed in several review articles, see for example the recent work by Giuliani et al. [10] and references therein.
Present experimental efforts do, however, not only concentrate on the production of new elements. Instead, the nuclear structure evolution in SHN is studied in detail, which is indispensable for answering fundamental questions about the nature of the strong interaction itself. Accurate experimental data on various nuclear properties such as, e.g., decay characteristics, energies of excited states, nuclear moments, the strength and location of shell effects, and many others provide insight into the underlying physics and challenge theoretical predictions, eventually leading to an improvement of the models.
Theoretical predictions of the heaviest nuclei were initially predominantly made by nuclear models using macroscopic-microscopic approaches in which the macroscopic description is based on the nuclear drop model [11] and shell-correction energies are introduced by an approach proposed by Strutinsky [12]. All models of this type predict the location of spherical proton and neutron shell closures to occur at Z = 114 and N = 184, see e.g. Ref. [13].
Self-consistent models using energy-density functionals, e.g., Skyrme-Hartree-Fock-Bogoliubov (SHFB) models and relativistic mean-field models (RMF) [14,15], have been applied to SHE since the late 1990s. Most of these models predict a spherical proton shell closure at Z = 120, but other functionals also obtain it at Z = 114 (SkI4), or even at Z = 126 (SkP, SkM*). There is generally a better agreement among the models with respect to the location of the spherical neutron shell closure that most of them predict to occur at N = 184, except for a number of RMF models that obtain a shell closure at N = 172.
Several authors have discussed how the location of these shell closures depends on details in the description of the underlying nuclear forces, for example on the values of the effective mass m * and the strength of the spin-orbit interaction. This has been discussed, for instance, by Rutz et al. [14] and by Bender et al. [16]. It has also been emphasized [16] that for self-consistent models using energy-density functionals the energy gap between the spin-orbit partners 2 f 5/2 and 2 f 7/2 may result in a shift of the maximum proton shell gap from Z = 114 to Z = 120.
The experimental strategy to benchmark these different theoretical models and to constrain some of their parameters is to obtain many different nuclear structure observables of nuclei in this region to provide information on the shell structure and the deformation. This is for example accomplished by systematic studies of nuclear masses, changes in mean square charge radii, and the measurement of the energies of excited nuclear states.
The low production rates of SHN and the resulting low number of events remain a major challenge in all experimental investigations. Still today, the only viable production scheme for superheavy elements are complete fusion-evaporation reactions. The resulting cross sections for evaporation residues are on the order of tens of nanobarn for the production of Rf isotopes and drop steeply to a level of one picobarn for Og. Even for high primary-beam intensities of > 6 × 10 12 particles per second as they are routinely delivered by the Universal Linear Accelerator (UNILAC) of GSI the corresponding yield for experiments is only in the range of few particles per minute or even some particles per week.
Therefore, a prerequisite for all experiments investigating these rare nuclei is an electromagnetic recoil separator providing a high suppression of primary beam-related particles while at the same time featuring a high transmission for evaporation residues. Consequently, the velocity filter SHIP [17] was one of the first experimental installations at the GSI to enable a superheavy-element-research program. The kinematic separation with SHIP is briefly described in Sect. 2.1. The design of SHIP and the experiments performed in the first 25 years of its operation that also led to the discovery of six superheavy elements with Z = 107-112 have been discussed in several previous reviews, see for example the review by Hofmann and Münzenberg [6,18], and are not subject of this article.
The experimental capabilities for a comprehensive SHE research program addressing the atomic and nuclear structure of the heaviest nuclides in the region Z = 100-110 around the deformed neutron shell closures at N = 152 and N = 162 were expanded in recent years. Nuclides investigated in recent experiments at SHIP as discussed in this article are depicted in Fig. 1. High-precision mass spectrometry and laser spectroscopy were introduced to the region of the heaviest elements at SHIP. These two complementary techniques offer new possibilities to observe the nuclear structure evolution, for example by studying shell closures via masses, by identifying long-lived isomers by mass spectrometry, and to determine nuclear spins and electromagnetic moments with laser spectroscopy. As already mentioned earlier, with these studies we can obtain a more complete picture of the structure of the heaviest nuclides.
To carry out this research program, we operate a suite of versatile instrumentation comprising the COMPASS detection system for α − γ spectroscopy in the focal plane of SHIP [19,20], the Penning-trap mass spectrometer SHIPTRAP for direct high-precision mass measurements [21], and the RAdiation Detected Resonant Ionization Spectroscopy (RADRIS) setup [22,23] for laser spectroscopy. Here, we review selected recent experiments on neutron-deficient isotopes of the elements No-Db carried out with the above-mentioned techniques in recent years. Section 2.2 gives a brief Besides the isotopes from the early element discovery campaigns at SHIP, the nuclides studied at SHIP in experiments with decay spectroscopy, mass measurements and laser spectroscopy, respectively, are indicated by the colored triangles. Open symbols indicated nuclides where the data analysis is ongoing and a publication is in preparation introduction to the nuclide production utilizing fusion-evaporation reactions. Mass measurements with SHIPTRAP and a discussion on some results are presented in Sect. 2.3. Section 2.4 presents the basics principles of laser spectroscopy together with the results of the first laser spectroscopy campaign in transfermium elements. In Sect. 2.5 the determination of nuclear properties from α-γ spectroscopy data is discussed, while Sect. 3 provides a brief summary and some conclusions.

The velocity filter SHIP
The two-stage velocity filter SHIP [17] (Separator for Heavy Ion reaction Products) has been the workhorse in Superheavy Element (SHE) research at GSI for about 45 years. The setup is schematically shown in Fig. 2.
It was designed for the investigation of products from complete-fusion reactions, so-called evaporation residues (ER), induced by heavy projectiles ( A > 40). Main design goals of the construction were to obtain a high transmission for the evaporation residues and an effective suppression of the projectile beam. The basic principle of the separation method is to use the velocity difference between projectiles and ER, which is given by Here v CN , v ER , v p denote the velocities of the compound nucleus (CN), the ER and the projectile (p), respectively. m p , m t , m CN are the masses of the projectile, the target Fig. 2 The present setup of the velocity-filter SHIP at GSI Darmstadt. Separation of recoils (yellow trajectories) from the primary beam (red trajectories) is achieved by their velocity through an arrangement of crossed electric fields, produced with two pairs of condenser plates (yellow blocks), and magnetic fields created by dipole magnets (blue). Two magnetic quadrupole tripletts (purple) shape the ion beam and a final dipole magnet (blue) improves the separation from scattered particles before the recoils reach the Time-of-Flight detectors (green) and the detector array. The scale on the left indicates the length in meters.
Reprinted with permission from Block [21] (t) nucleus and the compound nucleus. Evidently the ER velocity is always lower than the projectile velocity. The force F acting on an ion of charge q moving in a combination of crossed homogeneous magnetic (B) and electric (E) fields, arranged to deflect in opposite directions, is described by This effective force vanishes for ions with a velocity of v = (E/B) so that they will be transmitted through the combination of electromagnetic fields, while ions of any other velocity are deflected. Due to the reaction kinematics the distribution of the ER is peaked in forward direction and therefore SHIP was aligned in direction of the primary beam. It was believed at the time of construction that SHE would be preferentially produced in rather symmetric reactions, where ER are sharply focused in forward direction. Therefore, a small entrance aperture of ± 1.5 • in axial and radial direction was chosen to effectively suppress products from transfer reactions, which have a larger angular distribution. The technical layout was characterized by the following requirements: (a) Complete separation of the projectile beam from the ER (b) Collecting nearly all charge states of the ER (c) Accepting at best the complete range of the ER velocity distributions (d) High transmission (e) Stigmatic focussing (f) Velocity dispersion at the exit (g) Technical feasibility, specifically the operation of high-voltage condensers up to 600 kV (≈ 4.0 MV/m) that was technically challenging at the time of the construction. Taking all these requirements into account the separator was designed as a two stage velocity filter with separated electric and magnetic fields in the arrangement: quadrupole triplet-HV condenser-four dipole magnets-HV condenser-quadrupole triplet. SHIP accepts a relative velocity width of ±5% and a charge-state width of ± 10%. The HV condenser was chosen as the first dispersive element for technical reasons. As projectiles always have larger radii of curvature they are less deflected. So, even in case of a HV breakdown the projectiles will not hit the condenser plates, which could lead to severe damages when high beam intensities of, e.g., 1 pµA (6 × 10 12 particles per second) are used. The dipole magnets are of C-type with parallel entrance and exit field boundaries. So it is possible to vary all deflection angles and also the dispersion from zero to the maximum value. For experiments needing deflection voltages exceeding the maximum attainable, this is an important option. 1 The separated ions appear to pass undeflected through the field arrangement as long as the deflection angles are small. As the deflection fields of the magnets are only weakly focussing each stage was complemented with a quadrupole triplet.
The first filter stage is followed by a velocity slit, where an intermediate focus is produced by the first quadrupole triplet and the ER are already spatially separated from the projectile beam. The components of the second stage of the filter are arranged symmetrically to those of the first one with the plane of the velocity slit as the mirror plane and are operated anti-symmetrically, so that the velocity dispersions in both stages are added together [24].
After it turned out that rather symmetric reactions, e.g., 136 Xe + 170 Er → 306 122 * , are not the successful way to produce SHE [25] the strategy of 'cold fusion' reactions [26] using targets around doubly magic 208 Pb and medium-heavy projectiles ( 54 Cr, 58 Fe) to synthesize the new elements with Z = (107-109) was followed. Socalled cold-fusion reactions comprise combinations of heavy-ions beams with Pb and Bi targets resulting in a low excitation energy of the compound nucleus on the order of 10-20 MeV. For this excitation energy range, besides prompt fission of the compound nucleus, the de-excitation by one to two neutrons has the highest probability. It turned out that for these less symmetric reactions with a larger angular distribution of the ER, the small entrance aperture was a bottle neck, limiting the SHIP transmission to ≤ 30%. To proceed to heavier elements thus an increase of the transmission was necessary. The latter could be achieved by moving the target position as close as possible to the entrance of the first quadrupole. An increase by a factor of roughly two could be expected [27]. The prize to pay, however, was increased background of scattered projectiles passing SHIP with the ER-velocity. A sufficient suppression could be achieved by adding a dipole magnet of weak dispersion behind the second quadrupole triplet [28]. A deflection of 7.5 0 resulted to be a good compromise to remove unwanted particles and to avoid a significant transmission loss for ER. This configuration was since then adopted as standard operation mode.
The separation time is given by the flight-time through SHIP, which strongly depends on the mass ratio between projectile and target nuclei; for 254 No from the reaction 208 Pb( 48 Ca,2n) 254 No one obtains a value of T flight ≈2.3 µs, while for 270 Ds from the reaction 207 Pb( 64 Ni,n) 270 Ds a significantly lower value of T flight ≈1.4 µs is obtained.
In cold-fusion reactions for synthesis of SHE usually lead or bismuth are used as target material. The melting points of 600.61 K for lead and 544.4 K for bismuth render some difficulties in their use when high beam currents on the order of 10 12 -10 13 particles are required. Due to the deposited heat from the primary beam, fixed targets made of 'simple' foils will be destroyed immediately. Therefore, a target wheel rotating synchronously to the beam-pulse structure (5 ms, 50 Hz repetition rate) has been developed [29]. Target quality, size and performance have been continuously improved. In the present version [30] eight target segments are mounted on a wheel of 356 mm outer diameter and an 'effective' diameter of 310 mm with respect to the beam-spot position, rotating at a speed of 1125 rpm. The individual target segments that are mounted on such a wheel have a length of 120 mm and a height of 26 mm. 2 The material is evaporated or sputtered on a carbon layer of typically 40 µg/cm 2 and covered with a carbon layer of typically 10 µg/cm 2 to improve radiative cooling and to prevent sputtering of target material during irradiation; typical thicknesses of the lead or bismuth layers are ≈500 µg/cm 2 . The use of metallic lead or bismuth limited the beam intensity, so the use of compounds was considered. In particular, PbS (1397 K melting point) and Bi 2 O 3 (1100 K melting point) were identified as good choices [31]. To obtain homogeneous layers, the carbon-layer foil is heated during the production process to about 570 K. PbS targets turned out to be very robust and were irradiated with 40 Ar ions up to 3 pµA (1.86×10 13 /s); Bi 2 O 3 targets are less stable, they could be irradiated with 40 Ar ions up to 1.5 pµA.
The target quality is controlled on-line [32], using a 20 keV electron beam and measuring the attenuation of the current after passing the targets. Changes of structure can be resolved at a resolution of < 0.5 mm at a target velocity of 20 m/s. Sweeping the electron beam in radial direction enables to detect pinholes, changes of the effective thickness etc. over the whole target area.

Production of the nuclides of interest by fusion-evaporation reactions and separation from the primary beam by SHIP
The method of complete-fusion reactions, i.e. the total amalgamation of projectile and target nuclei has been so far the only successful way to produce isotopes of heaviest elements. For such complete-fusion reactions the energy relation depends on the kinetic energy of the projectile E p and is given by Reprinted by Heßberger [34]. The symbols represent experimental data taken from [33,34] while the lines are results from calculations using the HIVAP code. The HIVAP code has been described in Ref. [35].
Here, E CM is the kinetic energy in the center-of-system, Q = (m p + m t − m CN )c 2 is the Q value of the reaction and E * the excitation energy of the CN.
As both, projectile and target nuclei carry a positive electric charge, a minimum energy has to be applied to overcome the Coulomb barrier and bring both into contact. It is usually defined as the value where the probability for complete fusion is 0.5, which refers to a half density overlap of both reaction partners and is denoted as fusion barrier. For the production of superheavy nuclei using medium-heavy projectiles such as 50 Ti, 54 Cr, 58 Fe etc. and heavy target nuclei such as 208 Pb, 238 U, 248 Cm, etc., this simple description is not valid anymore.
At energies around the fusion barrier there is a strong competition between complete amalgamation of target and projectile nuclei resulting in a fully equilibrated mono-nucleus (compound nucleus) and binary processes leading to a strong exchange of nucleons or at best to a very short-lived mono-nucleus that reseparates already before mass and charge equilibrium has been reached. This is referred to as quasi fission. This reduction of the fusion probability is often denoted as fusion hindrance. In addition, superheavy nuclei are purely shell stabilized as liquid-drop fission barriers have vanished. Shell effects, however, are washed out at increasing excitation energies, and thus the survival probability of the compound nuclei drops rapidly at increasing excitation energies.
As a consequence there is only a narrow energy window favorable for the production of isotopes of the heaviest elements. For detailed investigations it is thus crucial to have information on the optimum energy at which the production cross sections are maximal. This is obtained measuring the production rates as function of the bombard-ing energy (or excitation energy) over a wide range. The result for such a measurement is shown in Fig. 3 for the so-called cold-fusion reaction 209 Bi( 48 Ca,xn) 257−x Lr. Data are taken from [33,34].
Evidently, the maximum cross-section is obtained for the 2n-channel ( 255 Lr) while it is dropping by about an order of magnitude for the 1n-channel ( 256 Lr) due to a decrease of the fusion probability towards lower energies, and the 3n-channel ( 254 Lr) due to increasing fission probability of the CN, a further steep decrease by a factor of about 40 is found for the 4n-channel. The lines are the results of calculations using the HIVAP code [35]. Parameters (sub-barrier fusion probability, fission barrier) were here adopted to reproduce the 1n-and 2n-cross sections. Evidently the trend of decreasing cross sections for increasing excitation energies is reproduced reasonably well.
It should be noted, however, that 48 Ca + 209 Bi is still a 'favorable' case. For heavier systems (in cold-fusion reactions), despite of lower fusion probabilities, 1n-channels dominate and energy regions for optimum production become even smaller.
Somewhat different conditions are found for 'hot' fusion reactions, used for the synthesis of elements Z = 114-118, using 48 Ca-projectiles and actinide targets in the range Z = 94-98, where due to lower 'fusion hindrance' maxima of the cross sections are found at higher excitation energies (E * ≈ 30-40 MeV) and excitation functions are broader. For details we refer to [8].

Mass spectrometry with SHIPTRAP
Penning-trap mass spectrometry (PTMS) is a well-established technique for nuclear structure studies. It relies on the direct connection and the absolute nuclear binding energy that can be obtained via accurate measurements of the atomic mass. Thus, systematic trends of masses, for example along isotopic chains, reveal signatures of nuclear shell structure and the onset of deformation [36]. High-precision mass measurements also provide a powerful method to identify (long-lived) isomeric states and pin down their excitation energy accurately as shown in several cases [37][38][39][40][41]. PTMS is suited for investigating long-lived isomers with half-lives of at least fifty milliseconds, a limit given by the time required for ion preparation. Due to the relation of the achievable mass resolving power with the available measurement time there is a tradeoff between highest resolution and shortest half-life to be made. However, for half-lives on the order of half a second and beyond the identification of states with low excitation energies is straightforward. This is of great interest in heavy nuclei that predominantly decay by internal conversion and for the identification of α-decaying isomers with similar half-lives of the isomeric state and the ground state. Besides revealing the nuclear structure evolution, direct mass measurements in the heaviest elements provide anchor points to fix the decay chains in the mass surface. Mass spectrometry techniques can also support the identification of (new) nuclides based on their mass-to-charge ratio, provided a sufficiently high mass resolving power can be achieved. This is specifically valuable if the decay modes or the half-life of the nuclide prevents the observation of correlated decay events. Combining the results from direct mass measurements with Q values from decay spectroscopy enables us to improve the mass values of heavier members of a decay chain. These may be inaccessible to direct mass spectrometry either due to half-life limitations or due to low yields. However, this requires that the decay schemes are well known. This is usually the case in even-even nuclides where the strongest α decay occurs from the ground state of the mother nucleus to the ground state in the daughter. In odd-odd and odd-A nuclei the strongest α-decay branch connects states with the same configuration, which is typically an excited state in the daughter. Here, a mass measurement can provide an independent result for the Q value.
PTMS is well described in the literature, see for example the reviews on PTMS of radionuclides by Blaum [42] and the review by Kluge [43]. A more general discussion on precision measurements of radionuclides has been presented by Blaum et al. [44]. The status of PTMS in the region of the heaviest elements has been recently discussed by Block [21,45]. Thus, here only a very brief introduction will be given with emphasis on aspects that are peculiar for mass measurements of the heaviest elements.
The installation of SHIPTRAP, a Penning trap behind SHIP, was proposed more than 20 years ago to probe the structure of heavy and superheavy nuclei via direct mass measurements [46]. This proposal was based on emerging techniques that paved the way for the efficient preparation of radionuclides from fusion-evaporation reactions as brilliant low-energy bunched beams facilitating an efficient injection into a Penning trap. One of the prerequisites was the buffer-gas-stopping technique to slow down particle beams with tens of MeV kinetic energy in an inert buffer-gas environment at pressures of around 50 mbar in less than one meter length [47][48][49]. The second ingredient was the application of radio-frequency quadrupole (RFQ) structures serving as cooler-buncher devices. This approach had been pioneered at ISOLTRAP [50]. It enables the cooling, stacking, and bunching of ions in helium buffer gas at moderate density (typical pressures are on the order of 10 −3 mbar). In this way, radioactive ions can be accumulated and prepared in parallel to an ongoing measurement in a Penning trap.
The schematic layout of SHIPTRAP is presented in Fig. 4. The first part is the second-generation buffer-gas stopping cell that is operated at about 45 K [49] for maximum cleanliness. The reaction products enter this cell after separation from the primary beam by SHIP through a thin-foil entrance window. This is made out of a low-Z material, typically titanium, and its thickness is chosen for each experiment according to the stopping range of the fusion products of interest. The typical thickness for Ti foils is on the order of 3 µm. Additional degrader foils may be inserted in front of the window to adapt the range. The window diameter of about 60 mm is adapted to the size of the spatial distribution of the reaction products at this position. The ions are then stopped in ultra-pure helium gas at about 50-70 mbar (room-temperature equivalent). The stopped ions remain in a charge state of 1+, 2+, or 3+ depending on their ionization potential [51]. They are extracted with electric fields that ensure the fast and efficient transport of the ions towards the exit. A funnel structure with a superimposed RF field prevents the ions from hitting the walls and focuses them onto a de-Laval nozzle of 0.6 mm diameter from which they are extracted in a supersonic jet of the buffer gas. A detailed description of the gas cell is given elsewhere [48,49].
The extracted ions enter an RFQ structure operated as ion guide transporting them to an RFQ cooler buncher. This consists of four axially segmented rods to form a potential gradient along the RFQ with a potential well at the end. The ions are cooled within few milliseconds in collisions with helium buffer-gas atoms and are accumulated in the potential well for a variable time adapted to the measurement cycle in the Penning traps. Ions can be stacked in the RFQ over times up to seconds. The ions are extracted as short bunches with low energy spread. The bunch length and the energy spread can be adapted to the requirements of an efficient capturing scheme in the Penning trap by varying the operation conditions of the cooler buncher.
A set of electrostatic ion-optical elements comprising einzel lenses and deflectors forms the transport section providing the ion bunches to the Penning-trap system. Two cylindrical traps are installed in the room-temperature bore of a superconducting 7 T-solenoid that features two homogenous regions. This double-trap concept was established at SHIPTRAP and is also employed at TRIGA-TRAP [52] and JYFLTRAP [53], for example. The installation of two Penning traps in one solenoid is cost efficient and simplifies the ion transport between the two traps. The first trap, the so-called purification trap, is used to remove unwanted reaction products and cool the ions of interest employing the well-established buffer-gas-cooling technique [54]. Depending on the experimental conditions such as the excitation amplitudes and the cycle time, mass resolving powers of m/ m ≈ 100,000 are achieved. This is usually sufficient to separate nuclear isobars. The separation of isobars is important to assure a pure ion sample for the following precision measurement. The presence of nuclear isobars is less of an issue for the heaviest nuclei produced by rather selective fusion-evaporation reactions where the production of isobars is usually suppressed due to the narrow excitation functions. Isobars may be 'produced' in electron-capture decays of the ion of interest during the storage in the trap. This can be detected in the measurement trap due to the high resolving power. An exception are isobars that may be produced during the storage and preparation of ions in the RFQ buncher and the purification trap by electron-capture decay for nuclides with half-lives on the order of the typical cycle times. The preparation of pure and well-cooled samples is also one of the prerequisites for high-precision measurement of the cyclotron frequency. For example, the cooling of the ions of interest in the first trap assures a minimal axial energy spread and a close-to-zero magnetron radius.
The measurement trap is connected to the purification trap by an orifice with 1 mm diameter and 51 mm length that acts as pumping barrier. A mass measurement is performed based on the determination of the trapped ion's cyclotron frequency ν c with the phase-imaging ion-cyclotron-resonance (PI-ICR) technique. This powerful method was introduced at SHIPTRAP a few years ago [55,56]. The cyclotron frequency of a trapped ion is obtained from a measurement of the phase of its radial motion. To this end, an excitation to a certain radius is performed applying an external RF field in a suitable geometry, followed by a free evolution for a defined (trapping) time until the extraction from the trap. To measure the phase, the radial ion motion is imaged on a position-sensitive ion detector. At SHIPTRAP, a commercial micro-channel plate detector with delay-line anode is used. A detailed explanation of the method and relevant systematic uncertainties has been given previously [55,56]. The PI-ICR approach has meanwhile become the new standard in PTMS of radionuclides due to its supreme performance with extremely high mass resolving power, a lower number of ions required to obtain a mass value compared to the former standard, the time-of-flight ion-cyclotron-resonance (ToF-ICR) method [57] and no need for an excitation-frequency scan. Low statistical mass uncertainties of a few 10 −9 are achieved for radionuclides as demonstrated in several experiments, in particular for mass difference measurements and for measurements of mass doublets [39,40,58]. Compared to ToF-ICR as used in the first mass measurements on nobelium and lawrencium isotopes at SHIPTRAP, PI-ICR provides a forty times higher mass resolving power for the same measurement time [55]. This is of particular importance for measurements of heavy nuclides and allows the identification of low-lying nuclear isomers. Note that a relative mass uncertainty of 5 keV for mass A = 250 corresponds to a relative precision of δm/m = 2 × 10 −8 . A gain in precision in PTMS can also be achieved by using higher charge states. However, charge breeding in an electron-beam ion trap, as for example used at TITAN [59], is (presently) not an option for the low ion rates handled with SHIPTRAP. However, due to the high cleanliness of cryogenic gas-stopping cells one can often profit from doubly and, in some cases, triply charged ions for mass measurements.
One limitation in PTMS is given by the need of calibrating the magnetic field strength B at the position of the ion of interest with high accuracy. This is usually accomplished by a measurement of the cyclotron frequency ν c−ref of a reference ion with well-known mass before and after the measurement of the ion of interest. In addition, this reference ion should have a mass-to-charge ratio similar to the ion of interest to avoid systematic uncertainties related to inhomogeneities of the trapping fields [60]. However, this may become challenging in the region of the heaviest elements due to the limited number of heavy reference ions with well-known mass. This is mitigated as most of the ions can be measured in a 2+ charge state so that 133 Cs + is a suitable calibrant ion with a mass-to-charge ratio sufficiently close to limit mass-dependent systematic uncertainties.
Another challenge is related to the magnetic-field stability that may become an issue for measurements on nuclides with low yield as for the heaviest nuclides that results in measurement times of several hours or even a few days. This is only feasible if additional measures are taken to improve the temporal stability of the electric and magnetic trapping fields. An indirect stabilization of the magnetic field is achieved by regulating the magnet-bore temperature and the pressure of the magnet's cryostat as employed at SHIPTRAP for several years [61]. The system has been recently improved further to reach a typical stability of the bore temperature of better than 10 mK and a pressure stability of the magnet's (LHe) cryostat within 0.1 mbar. This minimizes the impact on the cyclotron-frequency measurements to a level that is below the typical statistical uncertainty.
Continuous efforts to boost the performance of SHIPTRAP led to an overall efficiency on the level of 5-10% [49]. This value is mainly limited by the detection efficiency of the position-sensitive detector of about 30% and the charge-state distribution of ions extracted from the cryogenic gas cell. For ions with a high ionization potential, additional losses may occur by charge exchange with residual impurities in the RFQ buncher and the Penning trap. Mass measurements with lowest ion rates are furthermore limited by background that may arise from detector dark counts and by ions created in the radioactive decay of the ions of interest. Under typical experimental conditions at SHIPTRAP individual measurements can be performed with detected ion rates on the order of one ion per day. If a lower precision suffices, for example if only a mass number determination was required, then even lower rates could be handled.

SHIPTRAP results
The on-line commissioning of the initial SHIPTRAP setup with the first-generation gas stopping cell [47] was carried out in 2004 with isotopes of rare earth elements [62]. The following measurement campaigns focused on nuclides along the pathway of the astrophysical rapid proton-capture process that are favorably produced in fusionevaporation reactions [63,64]. In this way, for example, the mass of the nuclide 85 Mo, with N = Z + 1 was measured directly. Later on, the region of proton emitters among the rare-earth elements with mass numbers A ≈ 150 was investigated. Direct mass measurements provided accurate values of the proton-separation energy and allowed us to determine the location of the proton drip line in Tm and Ho. It should be noted that not all proton-unbound nuclides with S p < 0 show proton emission in their ground states as the partial half-live of the β decay branch is often shorter than that of the proton decay [65]. The nuclide with the lowest cross section that had been measured in that campaign was the proton emitter 147 Tm with a cross section of about 100-200 µbarn.
The first SHIPTRAP measurement campaign in the region of the heaviest elements was performed in 2008 when the mass of several nobelium isotopes starting from 254 No were measured directly for the first time [66]. These measurements were extended up to 256 Lr [67]. The results obtained until 2012 have been discussed in previous reviews [21,45] and are only briefly summarized here. In recent years, the measurements have been further extended towards heavier and ever-more rare nuclides in on-line measurement campaigns within the FAIR phase-0 program following a relocation of the SHIPTRAP setup [68]. A major performance improvement results from the integration of the cryogenic stopping cell. The new arrangement allowed us to inject the heavy ions of interest along the symmetry axis of the gas cell, whereas in the previous configuration the secondary ion beam had to be injected perpendicular to the extraction axis limiting the overall efficiency of the SHIPTRAP setup. The higher efficiency of the gas cell and its improved cleanliness together with the PI-ICR technique increased the sensitivity for mass measurements with SHIPTRAP significantly. Improved mass values for previously studied nobelium and lawrencium isotopes will become available soon as well as first direct mass measurements of 251 No, 254 Lr, 257 Rf, and 258 Db. In several cases also low-lying isomeric states have been resolved. The data analysis of the recent campaigns is still ongoing and first publications of the new data are in preparation. The measurements demonstrated that masses of nuclides produced with cross sections on the level of few nanobarn can be measured with SHIPTRAP. The heavy nuclides in the transuranium region that have been measured with SHIPTRAP are shown in Fig. 5.
The SHIPTRAP results are complemented by mass measurements of lighter actinides with the TRIGA-TRAP Penning trap in Mainz [52]. These actinide isotopes are still available in weighable quantities facilitating off-line measurements with a laser-ablation ion source. In addition to the previously published TRIGA-TRAP results on Pu and Am isotopes [69], additional actinides are being investigated in an ongoing measurement campaign. These are displayed by the open magenta squares in Fig. 5. TRIGA-TRAP can presently measure nuclides for which samples containing at least 10 15 atoms for the laser-ablation ion source are available.
Mass spectrometry of heavy radionuclides are nowadays also performed by the the KEK group working at RIKEN/GARIS using a so-called multi-reflection time-offlight mass spectrometers [70,71]. These devices reach mass resolving powers up to about 600,000 by extending the flight path in the spectrometer by reflecting ion bunches multiple times between electrostatic mirrors. The flight times of ions in such devices are on the order of ten milliseconds and precision of about δm/m ≈ 5 × 10 −8 -10 −7 have been reached [71][72][73][74]. The fast measurement scheme gives access to short-lived nuclei but the reduced mass resolving power compared to PTMS is insufficient to resolve isomeric states with low excitation energies that often exist in the heaviest nuclei. In recent years the KEK group has performed mass measurements of Fm and Md isotope behind the GARIS separator at RIKEN [70] and recently reported a first measurement in 257 Db with limited statistics [75].
Taking the latest data from direct mass measurements and other indirect techniques as they were considered in the atomic mass evaluation (AME 2020) [76] the nuclearstructure evolution around the deformed neutron shell closure at N = 152 can be studied. The visibility of this nuclear closure is enhanced by using so-called mass filters, i.e., a specific binding-energy difference. A commonly used one is the twoneutron separation energy S 2n . The two-nucleon-separation energies are favorable to display the gross nuclear-structure evolution compared to the one-nucleon-separation energies that feature the odd-even staggering reflecting the pairing interaction. A plot of the two-neutron separation energies of selected elements from Fm to Ds for neutron numbers from N = 140-170 is shown in Fig. 6. The presented data include the earlier SHIPTRAP results of direct mass measurements of nobelium and lawrencium isotopes [66,67,77] but not the yet unpublished results of the latest campaigns.
The S 2n values decrease smoothly with increasing N as expected and show a steeper drop at neutron numbers N = 152 and N = 162 where deformed shell closures are known to occur [78]. The size of the kink at the shell closure is indicative of the size of the shell gap and can be better visualized by the so-called empirical (neutron/proton) shell gap δ 2n/2 p . However, also in Fig. 6 it is seen that the strength of the shell gap at N = 152 changes in different elements and is largest in Fm at Z = 100. It should be noted that the strength of the deformed shell closures at N = 152, 162 is much smaller than the typical value obtained for a strong spherical shell closure such as the one at N = 126, for example, where the corresponding values is about 8 MeV. The general observation is that in the heaviest nuclides there are no longer distinct magic numbers but rather regions of extended shell stabilization. This has been also studied with several theoretical models, see for example the discussion by Bender et al. [16].

SHIPTRAP outlook
In this section, perspectives for future mass measurements with SHIPTRAP are addressed. Based on the present efficiency level of a few percent, measurements on nuclides produced with cross sections on the nanobarn level are feasible with SHIP-TRAP. Future upgrades along with the availability of a new accelerator that will provide ten times higher primary-beam intensity are expected to push this limit to a level of 10-100 pb. This opens up a number of new perspectives for direct mass measurements on nuclides with higher Z . A selection of suitable nuclear reactions for producing such isotopes is summarized in Table 1.
The majority of the given reactions utilizes lead and bismuth targets and has already been used at SHIP in decay spectroscopy experiments. These reactions lead to the production of neutron-deficient isotopes that are often too short-lived to be accessible with SHIPTRAP. This concerns for example the region beyond Sg around the deformed shell closure at N = 162 that is mostly out of reach with SHIPTRAP, except for the long-lived Hs isotopes 269,270 Hs [84], which already require the use of Cm targets. Also, the production of other more neutron-rich isotopes that are typically longer-lived requires hot-fusion reactions with radioactive actinide targets. In the case of Hs, an option would be for example 26 Mg( 248 Cm,xn) 269,270 Hs with a cross section of only 3 picobarn for 270 Hs. However, there are some technical challenges that have to be overcome. In the past the only actinide targets that have been used in experiment at SHIP comprise 238 U and 248 Cm that were irradiated with beams of 48 Ca, 54 Cr, and 64 Ni in experiments on copernicium, livermorium and in the search for element Z = 120 a few years ago [85,86]. Due to the larger angular distribution and the design of SHIP as discussed in Sect. 2.1, the transmission through SHIP decreases with increasing asymmetry of the nuclear reaction and drops rather steep for a ratio of projectile-to-target mass below 0.1. For 48 Ca-induced reactions with 238 U and 248 Cm the SHIP transmission is still on the order of 15-30% but drops for more asymmetric reactions such as 22 Ne on 197 Au to only about 5%. A second challenge for SHIPTRAP experiments is the low recoil energy that does not allow the use of pinhole-free Ti foils as entrance window any longer due their thickness. Thinner entrance windows with sufficient mechanical stability, also when cooling the system down to 40 K, and not jeopardizing the high cleanliness need to be identified and tested.

Resonance ionization laser spectroscopy at SHIP
A complementary approach for studying atomic and nuclear properties of the heaviest elements produced at the SHIP velocity filter uses the technique of resonance ionization spectroscopy (RIS). Laser spectroscopy in general is a versatile tool to unveil the atomic structure of an element, which in turn can be used to access properties of the nucleus [87,88]. Advancing techniques developed at the University of Mainz [89] for the laser spectroscopy investigation of elements heavier than fermium was accomplished, allowing for both shedding light on the atomic and on the nuclear structure of these heavy nuclides. In this section, a short introduction to the technique of resonance ionization spectroscopy is followed by a summary of the pioneering laser spectroscopic experiments on nobelium (Z = 102) performed at GSI, Darmstadt. Finally, an outlook to future activities will be given.

Resonance ionization spectroscopy
Laser spectroscopy uses optical transitions in the electronic shell of an atom to elevate electrons into orbitals with a higher energy. In RIS, the excited electron is removed from the atom by applying an additional laser excitation step with high photon fluxes and a photon energy sufficient to excite the electron above the first ionization potential (IP) [90,91]. This photo-ionization of the atom finally results in an ion which can be manipulated by electric fields and detected more efficiently. With this technique, properties of atomic transitions such as excitation energies and lifetimes can be measured, characterizing the element's electronic structure. From a measurement of the hyperfine-structure (HFS) splittings for isotopes with a non-zero nuclear spin nuclear properties such as the nuclear spin, the magnetic dipole moment, and the quadrupole moment can be obtained [88]. A more general overview of different laser spectroscopy techniques applied to heavy elements is given in a recent review [92]. A few neutron-rich isotopes of the heavy actinide elements einsteinium (Es, Z = 99) and fermium (Fm, Z = 100) can still be produced in weighable amounts of few picograms to nanograms in nuclear reactors by successive neutron captures and delayed β decays [93]. This allows some off-line measurements, but due to the limited sample sizes the information on the atomic structure is still scarce [92,94,95]. For the heavier transfermium elements (Z > 100), no atomic information was available experimentally prior to the investigations in nobelium (No, Z = 102) at SHIP [96], as these elements can only be produced in single-atom-at-a-time quantities in fusion reactions using intense heavy-ion beams from accelerators. Due to a lack of stable isotopes and the low production rates, theoretical predictions of the atomic structure are essential to guide any experimental search for atomic levels. These calculations are complicated by strong electron-electron correlations in these heavy systems and by the relativistic effects arising from the strong Coulomb force at the nucleus [97]. Having said that, the accuracy of the applied theoretical atomic models can be probed by the experimental determination of atomic-level energies and other atomic properties.
Once an atomic transition is found, it can be studied for different isotopes of the same element [88]. The obtained resonance signal will feature an isotope shift arising from a change in the nuclear volume and mass, while the interaction of the angular momentum J of the electron with a non-zero nuclear spin I = 0 results in a HFS splitting given by This splitting results in levels with a total angular momentum F where the separation of the individual levels is determined by the hyperfine parameters A and B. These hyperfine parameters are linked to the interaction of the magnetic field induced by the electrons at the position of the nucleus, B e (0), with the nuclear magnetic moment μ and to the electric-field gradient at the nucleus, δ 2 V δz 2 0 , interacting with the spectroscopic quadrupole moment Q s of the nucleus. The electronic part is typically inferred from HFS measurements on reference isotopes with known nuclear moments enabling the determination of nuclear moments for exotic isotopes [88]. Unfortunately, this does not work for the heaviest elements where a reference isotope with an independently determined nuclear moment does not exist. Therefore, atomic calculations of B e (0) and δ 2 V δz 2 0 are required for the extraction of nuclear properties. Also, calculated mass shift and field-shift factors allow us to derive the changes of the mean-square charge radii from measured isotope shifts.

The RADRIS experiment at GSI
The ER from the fusion reaction have typical energies of several 10 MeV (e.g. about 37 MeV in the case of 254 No) with an energy spread of several MeV rendering any direct laser spectroscopy impossible. Stopping and thermalization of the fusion products in a buffer-gas environment is therefore mandatory before performing laser spectroscopy. At SHIP, the so-called Radiation Detected RIS (RADRIS) method was applied successfully for laser spectroscopy of the transfermium element nobelium (Z = 102). This technique has originally been developed at University of Mainz, based on the experience with on-line laser spectroscopy on fission isomers in americium (Am, Z = 95) [89]. Here, the fraction of the incoming radionuclides, which was neutralized during stopping in argon buffer gas, was probed by laser resonance ionization spectroscopy and detected by their radioactive decay.
As the main fraction of stopped particles remains in an ionic state, a metallic filament was introduced to use electric fields for the collection of ions and subsequent thermal reevaporation of neutral species [98]. In this configuration, the filament material is crucial as a thermal desorption in combination with a low surface-ion yield is essential to the experiment. This has therefore to be verified and optimized for each element under investigation. The development of the RADRIS technique was continued at GSI, Darmstadt [23,99,100] and tantalum was found to be a suitable filament material for nobelium in off-line studies with the homologue element ytterbium, which features a similar electron configuration. The RADRIS setup is sketched together with the measurement procedure in Fig. 7. The incoming fusion products are slowed down in a buffer-gas volume and the stopped ions are guided by an electric field to a filament, a 125-µm thick tantalum wire for the nobelium experiments.
After a collection time, the primary beam is stopped and the filament is heated (to around 1050 • C for nobelium) to release the captured ions as neutral atoms. The evaporated atoms are probed by laser light and in case the photon energy matches an optical transition converted to a photo-ion by the very intense second laser. Due to the buffer gas in the cell the atom velocity is reduced, confining the area to be illuminated with the laser light as well as enabling the use of laser light with high pulse energy, a pre-requisite for an efficient photo-ionization. The tunable laser light is provided by excimer-pumped dye lasers while the ionization laser light originates directly from an excimer laser. Both lasers have a repetition rate of 100 Hz, which is suitable for typical drift times of the atoms in the buffer gas on the order of few milliseconds. The created photo-ions are guided to a silicon detector where their characteristic α-decay energy is detected, which in addition to the high efficiency of the setup ensures a high selectivity in the detection. The RADRIS method can be applied to α-decaying nuclides with half-lives of about 1-10,000 s. Depending on the decay mode of the collected radionuclide, the decay daughters remain on the filament and become accessible for laser spectroscopy as well. This opens up the access to nuclides that are difficult or impossible to produce directly. Proof-of-principle experiments for this approach have already been performed at the GSI. For example, the electron capture decay of 255 Lr has been exploited to produce 255 No and measure its hyperfine structure with RADRIS.

Recent results from laser spectroscopy
The RADRIS setup was coupled to SHIP for the atomic level search in nobelium. At the time of the first experiment was started, different theoretical predictions existed [101][102][103][104][105]. They all predicted a strong 1 S 0 → 1 P 1 ground-state transition with an excitation energy of about 30,000 cm −1 . However, due to the uncertainty of these predictions, a span of more then 1200 cm −1 corresponding to more than 3,000 scan steps had to be covered. Therefore, several beamtimes were required until finally in 2016 a first optical transition in nobelium was observed and characterized [97]. Besides the determination of the excitation energy to 29,961.45 cm −1 the saturation of this transition, measured with varying laser intensities, revealed a transition strength of 4.2 · 10 8 s −1 . The energy as well as the transition strength match very well with the predictions for the 1 S 0 → 1 P 1 ground-state transition [102,105], indicating that the strong electron correlations can be calculated at a high precision for heavy systems with nearly closed atomic shells. The lifetime of the excited level was additionally probed by varying the time delay between the first laser excitation pulse and the ionizing, second laser excitation pulse. This measurement revealed a delayed ionization for delay times much longer than the expected lifetime of about 2 ns for the 1 P 1 state obtained from the transition strength. The same behavior was observed in off-line investigations using stable Yb isotopes. Varying the buffer gas pressure revealed that the ion rate for large delays between the two laser pulses strongly increased with an increasing buffer-gas pressure.
These observations match with a strong population transfer from quenching by gas collisions to lower, but close-by, atomic levels with a sufficiently long lifetime enabling a delayed ionization. Such states are known to exist from theoretical predictions of the 3 D states. This assumption was verified by addressing Rydberg levels which are only accessible from certain atomic levels giving a unique fingerprint [106].
In a similar arrangement, the shift of Rydberg levels in Yb in the presence of a buffer-gas environment was studied off-line quantifying the resulting shift in the extraction of the IP [107]. With these findings and the on-line measurements of Ryd- Fig. 8 Convergence of the observed Rydberg levels in nobelium. The curve indicates a best fit of a Rydberg series convergence to the data. The observed three series clearly converge to two different limits as a result of buffer-gas quenching. For details see text and ref. [108]. Reprinted figure with permission from Chhetri et al. [108] berg levels in nobelium, is was possible to extract the first ionization potential of nobelium to 6.62621(5) eV [108]. This value is in very good agreement with theoretical predictions [102] and also matches recent measurements with a lower accuracy from surface-ionization yields at different temperatures reported by Sato et al. [109]. The observed Rydberg levels belonged to distinct Rydberg series converging to different limits as shown in Fig. 8. This behavior enabled determining the energy difference of the optically excited level and the atomic state to which the electron population quenches to be 310 cm −1 , while comparison to literature suggest this level to have a 3 D 3 configuration [108].
Besides probing the atomic structure of nobelium, which revealed more than 30 atomic levels to date, different isotopes were investigated as shown in Fig. 9. In addition, the resonance of the investigated 1 S 0 → 1 P 1 ground-state transition was measured for the isotopes 252,253,254 No. The isotope shift of the individual resonances reflects the change in nuclear volume determined by the mean-square charge radius [88]. The resonance spectrum of the isotope 253 No furthermore shows a hyperfine structure splitting originating from the non-zero nuclear spin as discussed in Sect. 2.4. From the angular momentum of the atomic ground state J GS = 0, the excited state J ES = 1 and the nuclear spin of I 253-No = 9/2 a total of three atomic transitions are expected, which were not fully resolved in the measurements. Nonetheless, the fit to the data is in agreement with this spin assignment. To infer nuclear properties from the measured hyperfine splitting, the atomic properties have to be known, c.f. Eq. 4. As no reference isotope is available for a relative measurement, atomic theory pro- Fig. 9 Measured spectra of the 1 S 0 → 1 P 1 transition in the nobelium isotopes 252−254 No. In 252 No the laser was operated at a larger bandwidth for an improved efficiency. The solid lines show the best fit to the experimental data. Figure adapted from Ref. [110]. Reprinted figure with permission from Raeder et al. [110] vided the atomic coupling factors for the hyperfine structure as well as for the isotope shift, enabling the determination of nuclear moments and changes in the mean-square charge radii [110].

Future prospects for laser spectroscopy
Besides the already discussed isotopes 252,253,254 No, which have already been investigated experimentally, the isotopes 251,255 No are in reach with the RADRIS technique. Their half-lives and their production cross sections are summarized in Table 2. In this consideration only cold-fusion reactions with lead and bismuth targets have been considered. The production of more neutron-rich nobelium isotopes is possible, using more asymmetric reactions. As already discussed in Sect. 2.3.2, the SHIP transmission is very low for very asymmetric reactions of light projectiles (such as C, O) and actinide targets (such as Pu and Cm). For 255 No an indirect production via the electron capture (EC) branch of 255 Lr is beneficial as the direct production via the 1 nevaporation channel in the 48 Ca + 208 Pb reaction always produces the isotope 254 No in the 2 n-evaporation channel with a similar cross section [33]. Both isotopes, 255 No and 254 No, have very close-lying α-decay energies around 8.10 MeV [79], reducing the selectivity from decay detection in this particular case. Some experimental data for 255 No has already been taken and an estimate of the obtained effective cross-section is stated in Table 2 while the data is still under evaluation. The half-lives were taken from the Evaluated Nuclear Structure Data Files [79]. For the decay-assisted spectroscopy of 255 No, i.e., the production by electron-capture from directly produced 255 Lr, the effective (eff.) cross-section is estimated The second isotope in reach is 251 No with a cross section of about 30 nb, about a factor of ten lower than in the case of 252 No. The short half-life of only 0.8 s is at the edge for the application of the RADRIS technique as the sequential collection and evaporation cycle is prone to decay losses and a fast switching is mandatory. A low-lying isomer with slightly longer half-life (1.02 s [79]) additionally complicates the measurements.
Besides extending the isotopic chain on both sides of the N = 152 shell by measuring further nobelium isotopes, the concept of laser spectroscopy on decay-daughter nuclei can be used to access additional nuclides that are inaccessible (of unfavorable) in a direct production scheme. The α-decay of nobelium, e.g., leads to fermium isotopes that can be probed with laser spectroscopy in an adapted cycle allowing for breeding these nuclei on the filament. For fermium some atomic levels were already reported [94,114]. In this context, the scan for atomic levels becomes unnecessary. In this concept new isotopes and elements will become available with a reduced efficiency due to the applied cycle and losses from recoils pushed into the filament material.
An obvious next step is the application of the RADRIS techniques to search for atomic levels in the next heavier element, lawrencium (Lr, Z = 103). The cross sections for the production of different Lr isotopes are shown in Table 2. However, even the highest cross section for the production of 255 Lr is an order of magnitude lower compared to 254 No. Nevertheless, the determination of atomic properties in lawrencium is of particular interest as the atomic ground state configuration is predicted to be 7s 2 7p 2 P • 1/2 , and thus different from its iso-electronic homologue, lutetium, which has a 6s 2 5d 2 D 3/2 ground-state configuration. For the lawrencium atom several predictions for atomic levels are available [105,[115][116][117][118] predicting two suitable strong ground-state transitions located around 20 400 cm −1 and around 28 400 cm −1 in the visible and the ultraviolet spectral range, respectively.
As discussed in Sect. 2.4.2 the desorption of collected fusion products from the filament is a crucial step in the RADRIS technique. As the 7s-shell of the atomic shell is fully occupied in nobelium, the atomic spectrum of lawrencium features p-and d-electronic orbitals leading to a higher desorption temperature due to an increasing refractory character of the transition metals along the actinide series. The IP of lawrencium (E IP =4.96 eV) [109,119] is significantly lower compared to nobelium, giving rise to an increased surface ionization fraction. The desorption was therefore studied off-line in the homologue element, lutetium, in a dedicated setup as sketched in the left part of Fig. 10. This setup consisted of a vapor source for evaporating Lu onto the filament to be tested and thus, the collection and evaporation cycle can be simulated with a filament analogue to the one used in the on-line RADRIS technique [23]. Resonantly tuned laser light is overlapped with the cloud of evaporated atoms to probe the neutral fraction by resonant laser ionization. The ions are guided through a quadrupole mass filter for selection according to mass-to-charge ratio and detected using a channel electron multiplier (CEM) in single ion counting. A detailed description can be found in [120]. This setup allowed the study of the desorption behavior as well as to monitor the surface-ion contribution for different temperatures and materials. From the different materials tested, hafnium turned out to be the most suitable material featuring a surface-ion yield, which was reduced by a factor of at least ten compared to tantalum, but required a desorption temperature of 1800 • C. A problem arose from the mechanical stability of the filament at high temperatures. The typical geometry of a 125 µm-thick wire as used with tantalum in the nobelium spectroscopy experiment turned out to be mechanically unstable. This can be seen in the right part of Fig. 10. The upper filament is a 125 µm-thick Hf wire heated to the desorption temperature of Lu and clearly is deformed. The solution was a change in the filament geometry to a 25 µm thick and 1 mm wide strip geometry as shown in the lower filament position of Fig. 10. This strip filament was also heated to the desorption temperature of Lu and showed only a minor deformation not hampering the experiment. With Hf as filament material and the optimized geometry a level search in lawrencium with RADRIS is now feasible.

A gas-jet setup at GSI for improved spectral resolution
The spectral resolution in the RADRIS technique is ultimately limited by pressure broadening of the atomic lines in the buffer-gas environment, where the laser spectroscopy is performed. A new technique for improved spectral resolution proposed by Kudryavtsev et al. [121] is the laser ionization in a well-collimated gas jet effusing out of the gas cell used for stopping of evaporation residues. The lower temperatures and gas densities in a gas jet enable laser spectroscopy with improved resolution. The crucial part is the de Laval nozzle needed for forming a well-collimated gas jet. This is required as the gas velocities in such a jet are about 550 m/s for argon as buffer gas. Therefore, an extended interaction volume of the gas jet with the laser light is required in addition to a narrow-bandwidth and high-repetition-rate laser system [122] for maximum efficiency. The technique was successfully applied on-line to neutron-deficient actinium (Ac, Z = 89) isotopes at the Leuven ISOL (Isotope Separator OnLine) separator with a tenfold improved resolution compared to in-gas-cell laser spectroscopy [123]. A continuous improvement of the technique including an efficient and thorough nozzle characterization optimized for laser spectroscopy of the heaviest elements is carried out in Leuven [124][125][126][127].
A dedicated gas-jet setup for laser spectroscopy of the heaviest elements at GSI is presently commissioned [128]. It combines the neutralization on a heated filament from the RADRIS technique with the laser spectroscopy in the effusing gas jet. A schematic layout of the system is shown in Fig. 11. The incoming recoil ions are stopped in argon buffer gas and guided by electric fields to a heated filament. On this filament, the ions neutralize and neutral atoms are transported further by the gas flow to the exit nozzle. The converging-diverging nozzle features a de Laval shape forming a well-collimated gas jet for high-resolution laser spectroscopy. A high-repetition-rate (10 kHz) dye-laser system is used for resonant excitation and ionization. The created photo ions are extracted and transported by a radio frequency quadrupole (RFQ) ion guide which is bend by 90 • to enable a longitudinal access of the laser light to the gas jet. The ions are finally detected by a CEM, albeit a silicon detector for a radiation-assisted detection is feasible as well. The achievable resolution is projected to be well below 400 MHz (FWHM) constituting an improvement by about one order of magnitude compared to the RADRIS technique. Such a resolution will benefit future experiments in particular the extraction of nuclear spins and electromagnetic moments from the HFS splitting. In addition, the setup allows a continuous operation giving access to both, much longer-lived nuclei by ion counting, and also to shorter-lived nuclei by minimizing decay losses that occur in the RADRIS technique. One case that can be tackled with the new setup is the 8 − K -isomer in 254 No with a half-life of 266 ms [79]. A laser spectroscopy measurement of the magnetic moment will enable the unambiguous determination of the configuration of this isomer that is still not unambiguously assigned as there are two different (quasi-particle) configurations that can form an 8 − K -state in 254 No [129].

Alpha-Gamma spectroscopy
An atomic nucleus consists of a number of nucleons (protons, neutrons) which interact via a strong attractive short-range force and the long range disruptive Coulomb force. The stability of a nucleus is thus governed by the ratio of the strength of these basic interactions. This feature becomes specifically important in regions of 'extreme' proton to neutron ratios (very neutron deficient or neutron-rich nuclei) and for the heaviest nuclei. In these cases, subtle details in the description of the strong force become important and will not only determine the location and the strength of the next spherical proton and neutron shells above 208 Pb (Z = 82, N = 126), but also the properties and stability of the heaviest nuclei in general. Under these considerations an extensive program to investigate properties of heaviest nuclei was started at SHIP [82,83,112,[130][131][132][133][134][135][136][137][138][139][140]. The main objectives of these experiments were: (a) study of decay properties, i.e. α decay, electron capture (EC), spontaneous fission; (b) identification of excited states, measuring the excitation energies and determining their spins and parities; (c) investigation of isomeric states with specific emphasis on K -isomers. From previous studies it was known, that excited levels show some systematic trends (with respect to excitation energies, spins and parities) along the isotone lines in even-Z odd-mass nuclei and along the isotope lines in odd-Z odd-mass nuclei. This feature has been consequently followed in experiments at SHIP for two decades, with systematic studies of even-Z nuclei along the the N = 147 to N = 153 isotone lines and along the isotope lines in odd-mass nuclei with Z = 101 and to lower extent Z = 103, 105. Specific nuclei to be investigated, however, were mostly selected with respect to known or expected production cross sections to obtain a maximum of information within limited irradiation times.

Detector setup at SHIP
Up to 2014, a detector system consisting of a 16-strip Si-'stop detector' (size 80 × 35 mm 2 ) surrounded upstream by a 'box' of six Si-strip detectors of equal size as the 'stop detector' was used in the focal plane of SHIP [28]. The strips were position sensitive in vertical direction, which was achieved by charge division. To discriminate Fig. 12 Transmission reduction for 254 No residues due to scattering in the TOF foils signals from implantation e.g., ER, scattered projectiles or decays e.g., α particles, SF events and light particles e.g., protons, α particles, deuterons produced in nuclear reactions with the target or backing material that may pass SHIP, another Si -detector of the same type and size was mounted behind the 'stop detector'. In front of the detector box, three time-of-flight (ToF) detectors were installed [141]. Due to scattering of the ER in the TOF-detector foils causing losses in transmission, usually only two detectors were used. This influence is shown in Fig. 12 for the production of 254 No in the reaction 208 Pb( 48 Ca,2n) 254 No.
For γ spectroscopy, a Ge-clover detector was mounted in close geometry behind the 'stop detector'. Gamma rays emitted in prompt or delayed (within a time window t(particle-γ ) ≤ 5 µs) coincidence with particle registration in the 'stop' or 'box' detector were measured using a Ge detector. Two types of Ge detectors were used: either a clover detector of VEGA type [142] consisting of four crystals, each of 70 mm diameter and 140 mm length, which were shaped and assembled to form a block of (124 × 124 × 140) mm 3 , or a smaller Clover detector consisting of four crystals, each of (50-55) mm diameter and 70 mm length, shaped and assembled to form a block of (102×102×70) mm 3 ('SHIP Clover'). Time differences between particles and γ events within the above mentioned time window of 5 µs, determined by the data acquisition system, were measured using a TAC with a resolution of t = 1 µs (FWHM) for the 'prompt' peak.
Detector pulses fed to ADCs started a coincidence time of usually 5 µs; after that time the ADCs were blocked and conversion and readout started, taking typically about 15 µs, which defined the deadtime of the system, during which incoming signals were not recorded.
Although the detector set-up was operated successfully for about 20 years, two severe disadvantages became evident. First, the above mentioned deadtime did not allow for an (effective) spectroscopy of short-lives isotopes or nuclear states (isomers). Second, the lack in sensitivity of the method to determine the position by charge division. Signals for low-energy particles such as CE and escaping α particles became quite low, often smaller than the threshold of the ADCs, ranging between 20 mV (under optimum conditions, i.e. no electronic noise and detectors of high quality) and 200 mV, if detectors had already suffered from radiation damages. Thus, often no or only one position signal was recorded. In addition, the position resolution suffered from non-linearities of the amplifiers in the range of low signals (< 500 mV). The position resolution for low-energy events ('escape' α particles and CE) was ≥ 1 mm (FWHM), whereas for α particles in the range 6-10 MeV it was ≈ 0.3 mm (FWHM). Therefore the detector system was replaced by the fully digitized setup COMPASS that uses pixelized detectors and has been described in detail along with the results of first performance experiments elsewhere [19]. Here we will summarize the main features.
The arrangement consists of a double-sided silicon-strip detector (DSSD) of 60 × 60 mm 2 in the focal plane. It has on each side 60 strips either in vertical or horizontal direction. Four single-sided silicon strip detectors (SSSD) of 60 × 60 mm 2 with 32 strips each forming a semi-box are arranged in front of the focal plane detector. The frames serving as holders for the silicon detectors are connected to copper tubes used for cooling with ethanol reaching typical temperatures of 263-268 K. The arrangement is placed inside an aluminum end cap having on each side circular windows of 80 mm diameter and a thickness of 1.5 mm to minimize absorption of low-energy photons.
For signal processing three options are available: 1. a pure analogue solution: charge-sensitive preamplifiers, pulse-shaping and peaksensing ADCs 2. a digital solution: the preamplifier signals are fed to FEBEX3 digitizers [143], at the heart of which are pipelining ADCs with 12 bit (14 bit) data range and 60 MHz (50 MHz) sampling rate 3. the use of ASIC-APFEL (ASIC for PANDA Front-End ELectronics) chips for detector read-out [144] placed inside the vacuum chamber of the detector box.
For the present purpose, a version of the signal processing was produced providing two input channels equipped with two different output channels. One of the latter had a switchable amplification factor of 16/32 with respect to the other one. The width of the output signals, which were fed to FEBEX3 modules, was ≈ 300 ns. For energy extraction a fast trapezoidal filter is used [145].

Decay study of 258 Db
In recent experiments at SHIP investigations of 247 Md and 255,257 Rf were continued. However, special focus was given to a decay study of 258 Db to solve two long-standing problems. First of all, to prove the existence of two long-lived states in 258 Db and its αdecay daughter 254 Lr, both decaying by α emission and/or by electron capture (EC). On the other hand, a direct proof of the EC decay of 258 Db could be obtained by measuring the K X-rays emitted during the EC process in delayed coincidence with spontaneous fission of the EC-daughter 258 Rf. 258 Db was first observed in an irradiation of 209 Bi with 50 Ti in 1981 [146,147]. In this experiment besides α decay of 258 Db also spontaneous fission with the same halflife as the α activity (T 1/2 ≈ 4 s) was observed and attributed to the decay of the same isotope. As 258 Db is an odd-odd nucleus it was not expected to undergo spontaneous fission itself. It was rather assumed that this isotope undergoes EC decay and fission of the EC-daughter 258 Rf, a known SF activity of ≈ 11 ms half-life, was observed.
In an experiment performed in 1982, it was attempted to prove this assumption by measuring the K X-rays emitted during the conversion electron (CE)-emission in delayed coincidence with SF of 258 Rf. The result of this study is shown in the upper panel of Fig. 13 [148] which was obviously not conclusive and therefore not published. In 2014, a successful measurement was eventually performed [149] and the photon spectra measured in delayed coincidence with fission events are shown in the lower panel of Fig. 13. Part of the SF events were measured in delayed coincidence with CE, i.e., we observed event sequences starting with implantation of an evaporation residue, followed by detection of a CE and terminated by a SF event ((ER, 258 Db)-CE-SF ( 258 Rf)). It was realized that (a) the fission events in delayed coincidence with CE had a longer half-life (4.4 ± 1.0 s) than those without CE (3.6 ± 0.3 s); although not completely unambiguous due to overlapping error bars this was seen as a hint for the existence of two long-lived states in 258 Db already suggested by a previous α-decay study [82]; (b) EC decay of 258 Db also populated two short-lived isomeric states in 258 Rf with half-lives of 15 ± 10 µs and 2.4 +2. 4 −0.8 ms. The decay scheme is sketched in Fig. 14.
The existence of two long-lived states in 258 Db was finally proven by the α-γ -decay studies [81], which resulted also in the existence of two long-lived isomeric states in the daughter nucleus, 254 Lr, and the granddaughter, 250 Md. The isomeric state in 254 Lr was meanwhile also confirmed by direct mass measurements with SHIPTRAP that will be published in a forthcoming paper.

Decay study of 257 Rf
The isotope 257 Rf was first observed by Ghiorso et al. [150] in the reaction 249 Cf( 12 C,4n) 257 Rf. They observed a complex α spectrum with energies in the range (8.5-9.0) MeV and measured a half-life of T 1/2 = (4.5±1.0) s. A couple of years later the results of Ghiorso et al. were confirmed by Bemis et al. [151], using the same reaction. In this experiment, photons were additionally measured in prompt and in delayed coincidence with K X-rays. Four groups of photon events were observed, which fitted to the expected K α1 , K α2 , K β 1 , and K β 2 X-ray energies of nobelium. It was thus the first direct Z -identification of a transactinide element. As some of the X-ray events were observed in delayed coincidence with α particles, Bemis et al. concluded that (at least) part of α decays of 257 Rf populate an isomeric level in 253 No. They attributed a half-life of T 1/2 = (31.1±4.1) µs and an excitation energy E * ≈ 300 keV to the isomer.
At SHIP, this isotope was produced in two ways. A direct production was obtained via the reaction 208 Pb( 50 Ti,n) 257 Rf [152,153], while it can be studied by an indirect production from the α decay of 261 Sg, directly produced in the reaction  [83,153]. It was shown that the two lines of highest energy are missing in the indirect production. This led to the assumption of a low-lying isomeric state in 257 Rf with a tentative configuration 11/2 − [725], that was populated in the direct production, but not by decay of 261 Sg. In a follow-up study the energy of this isomeric state was settled at E * = 70 keV, while the isomeric state in 253 No with a tentative configuration 5/2 + [622] was located at E * = 167 keV [83].
Within the experiment on the decay properties of 258 Db also a relatively short irradiation (ca. 24 h) of 208 Pb with 50 Ti to produce 257 Rf was performed. As a new result we observed a γ transition of 577 keV (2 events) in coincidence with α particles of 8296 keV, which was attributed to the decay of the 11/2 − [725] isomer into a-so far not reported-7/2 − [743]-state in 253 No decaying into the 5/2 + [622] isomeric state [154].
Specific emphasis was laid on the investigation of the EC decay of 257 Rf. For this purpose also the data from a previous study [83] were reanalyzed to detect possible differences in the EC decay of the ground state and the isomeric state. It was shown that in the direct production, where about 2/3 of the observed α decays stem from the isomer, a 257 Lr line of 8878 keV was observed, as also in the decay of 261 Db [83,155], while in the production via decay of 261 Sg a significant line at 8811 keV was recorded [156]. A careful analysis resulted in half-lives of T 1/2 = 0.20 +0. 16 −0.06 s for the 8811 keV transition and T 1/2 = 1.24 +0. 85 −0.36 s for the 8878 keV one, which clearly proved the existence of two states in 257 Lr decaying by α emission. As the latter was observed in the direct production it was attributed to a high spin state, while the 8811 keV transition was attributed to the decay of a low spin state. This observation was not surprising as a similar behavior was encountered also in the lighter odd-mass isotopes 253 Lr [130] and 255 Lr [112,157]. The (partial) decay scheme for 257 Rf deduced from the recent SHIP experiments is shown in Fig. 15.

Isomeric states in 255 Rf
After identification of a K -isomeric state in 270 Ds living longer than the ground-state [158] and the first γ spectroscopy investigations of K -isomers in 254 No [129,159] investigation of K -isomers came into the focus of nuclear spectroscopy of SHE. At SHIP, new K -isomers or candidates for K -isomers were identified in 251 No [134], 252 No [135,160], 253 No [138,161], 255 No [137], 266 Hs [162], and 258 Rf [149].
Within the experiment on the investigation of 258 Db we also performed a short irradiation (≈ 40 h) of 207 Pb with 50 Ti to produce 255 Rf. Main focus was to measure the total kinetic energy (<TKE>) of SF (see next section), but also to search for K -isomers in that nucleus.
To detect events from the decay of possible isomeric states, correlations of the type ER (implantation signal)-CE (possibly in prompt coincidence with a γ event)-α decay / SF (from 255 Rf or 251 No) were searched for. But still one had to consider a possible population of the known 5/2 + isomeric state in 255 Rf (E * ≈135 keV, T 1/2 = 50±17 µs) [140]. This isomer is known to decay essentially by CE emission of E < 150 keV. So, being conservative, CE energies above 200 keV could be assigned to the decay of another isomeric state. Two activities were identified [80]: (a) one with a half-life of T 1/2 = 38 +12 −7 µs and CE energies essentially below 370 keV, and (b) one with a half-life of T 1/2 = 15 +6 −4 µs and CE energies above 370 keV. Based on the observation of three correlations of the type ER-CE1-CE2-α/SF the events with the longer half-life were attributed to an isomeric state at E * ≈ (1.15-1.45) MeV, the one with the lower half-life to a state at E * ≈ (0.9-1.2) MeV.
For some of the CE also γ events were observed in prompt coincidence, but no line structure was visible due to low statistics. Also no spin and parity values could be determined. This has to be left for further, more detailed studies.

Spontaneous fission properties of 255,256,258 Rf
Important features in spontaneous fission studies are the determination of the total energy release (<TKE>) and the mass distribution of the fission fragments. Implantation of nuclei into silicon detectors is, however, not a well-suited technique for such investigations: (a) due to the high ionization density of the heavy fission fragments, part of the created charge carriers will recombine and so there is no linear dependence between the height of the signal and the energy of the fission fragments; (b) the implantation depth of the ER (typically <10 µm) is lower than the range of the fission products (typically 20-25 µm). Therefore, recorded signals will be a mixture of the sum of the energy of both fragments (E(SF1) + E(SF2)) and the energy of one fragment and the energy loss of the other one (E(SF1) + E(SF2) or E(SF1) + E(SF2)). Despite of these restrictions the 50 Ti beam was used to obtain more precise information on (<TKE>) of 255,256 Rf. Quite precise measurements of (<TKE>) and the mass distribution have been reported for 258 Rf [163]. As the experimental conditions (detector system, calibration, implantation depth) were quite similar for 255 Rf (production: 207 Pb( 50 Ti,2n) 255 Rf), 256 Rf (production: 208 Pb( 50 Ti,2n) 256 Rf), and 258 Rf (production: 209 Bi( 50 Ti,1n) 258 Db EC → 258 Rf), the measurement of 258 Rf was used as a reference for improving the data for 255,256 Rf. In addition, results from a previous SF study of 252 No [85] were reanalyzed for comparison. Using (only) the events recorded as coincidences in the 'stop' and 'box' detectors (see Sect. 2.5.1) we obtained [80] (Previously reported values are given in brackets): The <TKE> distributions were found to be narrower than for 252 No, which is known to fission asymmetrically [164]. This is seen as a signature for (nearly) symmetric fission of the three investigated Rf isotopes, in accordance with the conclusions drawn for 256,258 Rf in [163].
The new results (red squares) are compared with published <TKE> values in Fig. 16; the lines refer to empirical systematics from Unik et al. [166] (dashed line) and Viola et al. [167] (full lines). In addition, the <TKE> distributions were analyzed with respect to possible contributions for 'bimodal' fission. Small effects for 255,256 Rf were indicated, but far from being unambiguous. Therefore such studies have to be left for further investigations using experimental techniques more suited for such cases.

Decay study of isotopes produced in the reaction 181 Ta( 48 Ca,xpyn) 229−xpyn Np
Nuclei in the vicinity of closed proton and neutron shells are of specific interest, as their decay properties reflect changes in the strength of the shells along isotone or isotope lines. From the results of a recent decay study of the uranium isotopes 222 U and 221 U that were identified for the first time, a weakening of the influence of the N = 126 shell on the decay properties of these uranium isotopes was concluded [168].
Thus, the investigation of the N = 126 shell strength towards elements with higher Z numbers is of interest. As a first step in this direction and also to test the performance of the new SHIP detector system COMPASS (see Sect. 2.5.1), we performed an investigation of nuclei produced in bombardments of 181 Ta with 48 Ca, specifically those starting from 225 Np (4n-channel), 226 Np (3n-channel), 222 Pa (α3n-channel), 225 U (p2n-channel), and 223 Pa (α2n) [20]. We also had the intension to obtain decay data of higher quality for 225,226 Np and possibly to identify the then unknown isotope 224 Np that has meanwhile been observed [169]. From a technical point of view, it was of interest that the α-decay chains of 225,226 Np include very short-lived members, 218 Ac (T 1/2 = 1.08 µs) in the case of 226 Np, 221 Pa (T 1/2 = 5.9 µs), and 217 Ac (T 1/2 = 69 ns) in the case of 225 Np, which appear as pile-up events in previous experiments using the implantation technique and analogue electronics, but were now measured with the fully-digitized detector system COMPASS.  [165] In general, published decay data were reproduced or improved. New decay data were obtained for 225 Np, 226 Np, and 222 Pa. For 226 Np and 222 Pa new high-energy α transitions of 8.18 ± 0.02 MeV ( 226 Np) and 8.63 ± 0.04 MeV ( 222 Pa) were observed that had not been reported before. For 225 Np α-decay energy and half-life were measured simultaneously for the first time. Andreev et al. [170] had reported only an energy of 8.63 MeV, while Devaraja et al. [171] gave a half-life value of ≈ 3.8 ms obtained from the time difference between the decays of 229 Am and 225 Np of 5.5 ms of one observed event, but they did not report an energy value, as the event was observed as a pile-up with the daughter nucleus 221 Ac. The SHIP experiment of Mistry et al. [20] delivered a somewhat higher α-decay energy of 8.8±0.1 MeV and on the basis of two observed events a half-life of 0.31 +0. 75 −0.13 ms, which is about an order of magnitude lower than the value given by Devaraja et al..

Summary and conclusions
In recent years, the experiments at SHIP focussed on detailed investigations of the nuclear structure of nuclei in the region Z = 102-105 in the vicinity of the deformed neutron shell closure at N = 152. Nuclides in this region can still be produced with rates that result in rather high statistics even in shorter experiments. We have performed systematic studies of different and often complementary nuclear properties by α-γ spectroscopy, by direct mass measurements, and by laser spectroscopy contributing to a better understanding of the nuclear structure evolution in this region. In particu-lar, we have studied the evolution of the shell closure at N = 152 in the elements Z = 102-105 investigating binding energies of 252−255 No and 255,256 Lr isotopes from directly measured nuclear masses [66,67], by the change in nuclear charge radii for 252−255 No by laser spectroscopy [110], and by the determination of excited nuclear states in 255,257,258 Rf and 257,258 Db isotopes by α-γ spectroscopy. We obtained more detailed information on the structure and the decay modes of these neutron-deficient Rf and Db isotopes. In particular, we identified the electron-capture-decay of 258 Db unambiguously by delayed coincidences with X-rays in the daughter. This approach will become more important in the future as EC decays play an increasingly important role in the assignment of decay chains originating from more neutron-rich superheavy nuclides. Also, the complex decay scheme of 258 Db was further elucidated and evidence for two long-lived isomeric states in 258 Db was revealed [81].
Accurate mass measurements of several No and Lr isotopes provided anchor points in the mass surface that allowed us to improve the masses of heavier nuclides that are linked by α-decay chains, for example 270 Ds [77]. The determination of groundstate masses of odd-odd and odd-A nuclides by mass spectrometry provide Q-values essential to establish unambiguous nuclear level schemes in conjunction with decay spectroscopy. In the future, the identification of isomeric states and the accurate determination of their excitation energy by mass spectrometry with SHIPTRAP will be exploited. Thanks to the powerful PI-ICR technique this is now possible for longlived isomers with lowest excitation energies of tens of keV. Additional synergies are expected from trap-assisted decay spectroscopy experiments for select nuclides where SHIPTRAP will be used as high-resolution mass separator to provide not only isotopically pure samples, but even nuclear-state selected samples for subsequent decay studies. The capabilities of this approach have already been demonstrated for 213 Ra [172].
Information on the deformation and shape of heavy nuclei can be obtained from the observation of rotational bands and from hyperfine laser spectroscopy and optical isotope shift measurements. Laser spectroscopy also provides information on nuclear spins, electromagnetic nuclear moments and changes in the mean-square charge radii [92]. Together with decay spectroscopy this gives a handle on nuclear configurations. Laser spectroscopy of nobelium isotopes has demonstrated that this method can be applied to the heaviest elements even if no experimental data on the atomic structure exists. From the first experiments information on the changes in the mean-square charge radii and the deformation in three No isotopes were achieved [110]. The results are in good agreement with theoretical calculations by energy-density functionals. The hyperfine spectroscopy of 253 No provided the nuclear dipole and quadrupole moments and confirmed the spin-parity assignment from earlier decay spectroscopy experiments.
The combination of three complementary approaches, decay spectroscopy, laser spectroscopy, and mass spectrometry, for very heavy nuclides is presently only available at SHIP at the GSI. One prominent example where this may help us to solve a long-standing puzzle is the structure of the 8 − K isomer in 254 No. This isomer has been investigated by decay spectroscopy several years ago but its structure is still debated [129,137,159,173]. Two quasi-particle configurations in 254 No can form an 8 − state, a quasi-neutron and one quasi-proton configuration. Despite several experi-mental attempts it has not been possible to unambiguously determine the configuration of this isomer yet. The answer could be revealed by high-resolution laser spectroscopy by a measurement of the magnetic moment. Such an experiment will be enabled in the near future by the new gas-jet setup discussed in Sect. 2.4.5.
The extension of our experiments to heavier nuclides will be accompanied by continuous developments of new methods and improved experimental setups. The PI-ICR method, the COMPASS detector and the gas-jet setup are examples of this successful strategy. Our experimental program will also benefit from future upgrades of the linear accelerator chain at the GSI [174] that are expected to result in an increase of primary-beam intensities by up to an order of magnitude. The use of hot-fusion reactions with radioactive actinide targets will be needed to extend the reach towards more neutron-rich nuclei. While the known-how in handling radioactive actinide targets exists at GSI as shown in several experiments as SHIP [86] and TASCA [175,177], some challenges remain. One limiting factor is the reduced transmission of SHIP for more asymmetric fusion reactions, which can only be overcome in a future upgrade.
In addition to the exploration of the region of the heaviest elements, the study of selected neutron-deficient isotopes of lighter nuclides is an area where experiments at SHIP remain competitive. The region around the Np isotopes studied in the COMPASS commissioning is one example. The search for exotic decay modes such as betadelayed fission [176] is another topic of interest.