Search for elements 119 and 120

A search for production of the superheavy elements with atomic numbers 119 and 120 was performed in the 50 Ti + 249 Bk and 50 Ti + 249 Cf fusion-evaporation reactions, respectively, at the gas-ﬁlled recoil separator TASCA at GSI Darmstadt, Germany. Over four months of irradiation, the 249 Bk target partially decayed into 249 Cf, which allowed for a simultaneous search for both elements. Neither was detected at cross-section sensitivity levels of 65 and 200 fb for the 50 Ti + 249 Bk and 50 Ti + 249 Cf reactions, respectively, at a midtarget beam energy of E lab = 281 . 5 MeV. The nonobservation of elements 119 and 120 is discussed within the concept of fusion-evaporation reactions including various theoretical predictions on the ﬁssion-barrier heights of superheavy nuclei in the region of the island of stability.


I. INTRODUCTION
To date, 118 chemical elements are known. They fill the periodic table of the elements until the end of the seventh row. The heaviest elements with proton numbers Z = 114-118 have been synthesized only in fusion reactions of the doubly magic 48 Ca (Z = 20) nucleus with nuclei of radioactive isotopes of actinide elements from plutonium (Z = 94) to californium (Z = 98) [1,2]. Elucidating the nuclear, atomic, and chemical properties of superheavy elements (SHEs) is † † Present address: Cyclotron Institute, Texas A&M University, College Station, Texas 77843, USA. a fundamental quest in chemistry and physics [2][3][4]. One of the main goals of SHE research is the search for an island of stability arising from the presence of closed nuclear shells, which are predicted to inhibit spontaneous fission (SF) of the superheavy nuclei (SHN).
Currently available experimental data on the decay properties of known superheavy nuclei [1,[5][6][7][8][9][10][11][12][13][14][15][16] indicate a stability against fission, thus confirming the concept of the island of stability, but to date the exact location of the center of the island of stability, i.e., proton and neutron shell closures, and its landscape are not yet known. For a long time it was assumed that Z = 114 and neutron number N = 184 would form closed shells [3]. However, current information from experimental data and modern theoretical calculations does not exclude that the next closed proton shell occurs at Z > 118 [4]. From chemical and atomic perspectives, SHEs beyond oganesson (Og, Z = 118) will start the eighth row in the periodic table. Data on SHE from the eighth row are of great interest for the verification of the periodicity of the elements and the influence of relativistic effects on chemical properties [2].
The synthesis of SHEs beyond Og faces, however, many experimental challenges. One of them is the need to use fusion-evaporation reactions with projectiles heavier than 48 Ca [17][18][19], because of insufficient amounts of materials of elements with appropriate proton numbers, Z > 98, to make a target [20]. With this constraint, the four different reactions 64 Ni + 238 U [21], 58 Fe + 244 Pu [22], 54 Cr + 248 Cm [23], and 50 Ti + 249 Cf [24] have already been examined for the synthesis of SHE with Z = 120. However, none of these experiments provide evidence for the synthesis of the new element.
The use of 50 Ti as a projectile and that of 249 Bk and 249 Cf as targets appear to be the most promising combinations for the synthesis of elements 119 and 120 [17][18][19]25]. 249 Bk, which decays by β − into 249 Cf with a half-life of only 327.2(3) d [26], is a unique target. Starting with a freshly prepared, pure 249 Bk target, the amount of 249 Bk will continuously decrease over time, and the amount of 249 Cf will increase [16]. This situation provides a unique opportunity to simultaneously search for a direct production of two SHEs in a single longlasting irradiation. This was the case in the bombardment of a 249 Bk target by 48 Ca, where the two elements tennessine (Ts, Z = 117) and Og were observed as evaporation residues (ERs) from the 48 Ca + 249 Bk and 48 Ca + 249 Cf reactions, respectively [27].
By exploiting this feature of the 249 Bk target material, we searched for the SHEs with Z = 119 and 120 in a four-month long experiment with a 50 Ti beam.

II. EXPERIMENTAL SETUP AND CONDITIONS
Isotopically pure 249 Bk target material was produced at Oak Ridge National Laboratory, USA [20], and a sample of 12 mg was shipped to Johannes Gutenberg Universität Mainz, Germany, for the target production. Four banana-shaped target segments with thicknesses of 0.37(4), 0.53 (5), 0.53 (5), and 0.50(5) mg/cm 2 were produced by molecular plating [28] on 0.99(5)-mg/cm 2 Ti-backing foils, each with an area of 6 cm 2 . The average areal density of the target was 0.48(5) mg/cm 2 . Because of the relatively short half-life of 249 Bk, it was essential to start the experiment as soon as possible to maximize the probability for the discovery of element 119. Target irradiation started within about one month after the target production [29].
The evolution of the areal densities of 249 Bk and 249 Cf in the target is shown in Fig. 1(a). One can see that the target with an average areal density of 0.48(5) mg/cm 2 consisted of ≈0.43 mg/cm 2 249 Bk at the beginning of the experiment. This converted slowly into 249 Cf. At the end of the experiment the ingrowth of 249 Cf reached ≈ 0.17 mg/cm 2 . The four target segments were mounted on a wheel, which rotates synchronously to the time structure of the beam. To ensure safe operation, all target segments were continuously monitored using different methods such as online temperature readings during the experiment with a pyrometer [16,30].
Prior to the experiment, a high-intensity and long-term stable 50 Ti beam was developed and established at GSI. A heavy-ion beam of 50 Ti with charge state 2 + was produced in a Penning ion source. These ions, further, were stripped to a charge state 12 + and accelerated by the Universal Linear Accelerator (UNILAC) in a pulsed mode with 5-ms pulse length and 50-Hz repetition frequency to an energy of 300 MeV. A total beam dose of 3.6 × 10 19 particles passing the target was accumulated in two experimental campaigns, which overall resulted in about four months of irradiation. The average intensity of the beam on the target was about 0.65 particle μA (4 × 10 12 particles/s). In Fig. 1(b), the chronology of beam intensity over the whole four-month period is shown. To maintain a high beam intensity over a long time period, fresh ion-source material was supplied daily.
Before impinging on the target, the beam passed through a 50-μg/cm 2 thin carbon foil mounted on a wheel on the same axis as the target wheel, with both wheels rotating synchronously. The carbon foil was used as a charge stripper for 50 Ti 12+ ions to ensure their safe deflection in the dipole magnet into the direction of the beam stop. The beam energy in the center of the target was estimated by using the SRIM code [31]. Energy losses of the initial beam in the carbon and titanium foils were directly calculated using the database of SRIM. For calculation of the energy loss in the actinide target, its most probable chemical composition in forms of 249 Bk 16 2 O 3 and 249 Cf 16 2 O 3 was used, and the proper ratio of 249 Bk to 249 Cf atoms in the target was taken into account. The average beam energy was calculated at E lab = 281.5 MeV in the center of the target (laboratory frame), which is not affected by variation of the fractions of 249 Bk and 249 Cf in the target. Beam energies entering and leaving the target are 283.9 and 279.1 MeV, respectively.
The excitation energies of the compound nuclei (CN) 299 119 * and 299 120 * , which could be produced in the fusion of 50 Ti with 249 Bk and 249 Cf, respectively, were estimated by using the known experimental mass excesses of projectile and target nuclei [32] and theoretical values from Ref. [33] for the CN. With the midtarget beam energy of E lab = 281.5 MeV, excitation energies of E * = 43.2 and 37.6 MeV result for 299 119 * and 299 120 * , respectively. At these excitation energies the 3n and/or 4n evaporation channels are expected to be predominant and corresponding heavy ERs should be observed [1,34]. The gas-filled recoil separator TASCA [12,35] was filled with helium gas at 0.8-mbar pressure. Its magnetic fields were set to guide heavy ions with a magnetic rigidity of Bρ = 2.16 Tm [36] to the center of the focal plane detector. This magnetic rigidity ensures a safe isolation of ERs from both fusion reactions from the primary beam and products of other reaction channels like elastic and (deep) inelastic scattering [37,38]. The efficiency of TASCA for the collection of ERs from the 50 Ti + 249 Bk/ 249 Cf reactions was estimated by performing Monte Carlo simulations [37]. An average transmission of 55% was derived for the applied experimental conditions and taking into account theoretical predictions on the shape of the ER excitation functions for the 50 Ti + 249 Bk/ 249 Cf reactions [34].
The ERs passing through TASCA entered the detector chamber and first passed through a multiwire proportional counter (MWPC). The anode signal was read out and stored in coincidence within about 5 μs to any event registered in the implantation detector.
The focal plane detection system (FPDS) of TASCA consisted of a double-sided silicon strip detector (DSSD)-based implantation detector (hereafter: stop detector), with eight DSSDs (hereafter: box detectors) mounted perpendicular in the backward hemisphere of the stop detector to form a five-sided box configuration. The stop detector consisted of 144 vertical (X ) and 48 horizontal (Y ) 1-mm strips on the front and back sides, respectively. The 144 vertical strips faced TASCA and had 0.1-mm interstrip pitch. The 48 horizontal strips provided the position information along the Y axis. Each box detector was 72 × 48 mm 2 in size and had 16 strips on each side, oriented perpendicular to each other. The longer strips were faced inside the box configuration. In the data processing, signals from every two neighboring strips of the box detectors were combined. Detailed descriptions of the FPDS are given in Refs. [39,40] Two adjacent single-sided Si-strip detectors having together the same size as the stop detector were mounted directly behind the stop detector to register particles passing through the stop detector (veto detector). The veto detector was used to discriminate real α events from low-energy sig-nals originating from light charged particles passing through the separator and the stop detector [9].
The Combined Analog and Digital (CANDI) [41] data acquisition (DAQ) system was used for processing the data collected with the FPDS. Signals from the front 144 vertical strips of the DSSD, box, veto, and MWPC detectors were processed in a standard way, i.e., preamplified, amplified and shaped, and digitized by using peak-sensing analog-to-digital converters (ADCs). All preamplified signals were duplicated. Spectroscopic amplifiers with two gains differing by a factor of 11 were used to create two branches for energies of α particles and for high-energetic particles, respectively. These signals were independently stored in the analog part of the CANDI, which had a dead time of ≈35 μs. Preamplified signals from the horizontal strips of the stop detector were digitized by 60-MHz-sampling ADCs by storing their shapes in 50-μs-long traces (8 μs before and 42 μs after the trigger). Finally, data streams from analog and digital DAQs connected to the FPDS were combined as single data, which allowed the determination of the time, spatial coordinates, beam-on/off status, energy, and shape of each detected event [42]. The advantages of CANDI-type systems for the solution of various physics and measurement technical issues and for superheavy element search experiments are demonstrated in Refs. [11][12][13]16,41,[43][44][45][46].
The efficiency for the detection of α particles with full energy emitted by nuclei implanted in the FPDS is estimated to be 76(4)%. The efficiency for the detection of fission events is 100%. The energy resolution (full width at half maximum) of individual strips of stop and box detectors prior to the experiment was ≈40 keV for 5.8-MeV α particles from an external α source placed in front of the DSSD. The final energy calibrations were done using the α decays of nuclei produced in a preparatory irradiation using the 50 Ti + 176 Yb reaction [41]. In case of the Y strips of the DSSD, the full energy of an event was sometimes shared between two neighboring strips while the energy from the front side was collected by a single X strip. On average, such split signals were observed in 16% of all cases throughout all Y strips. The data acquisition was triggered by any event registering more than about 600 keV 064602-3 in a front (X ) and/or more than about 500 keV in a back (Y ) strip of the stop detector.
Calibration of the high-energy branch was done with an external four α-line source [5]. With such a calibration, the measured energies of fission fragments from 256 Rf were distributed in the range of 50-200 MeV [44].

III. EXPERIMENTAL RESULTS: SEARCH FOR ELEMENTS 119 AND 120 IN THE CORRELATION ANALYSES
The low-energy spectrum (analog part of the data stream) of events registered during the 15-ms beam-off period after each 5-ms pulse is shown in Fig. 2. This spectrum clearly shows the peaks and pileup events corresponding to α decays of nonfusion products originating from both 50 Ti + 249 Bk and 50 Ti + 249 Cf reactions. A detailed investigation of such nonfusion products from the 50 Ti + 249 Cf reaction obtained at TASCA was carried out in Ref. [43]. The nuclei produced in the present experiment as products of nonfusion reactions were similar to this 50 Ti + 249 Cf study [43].
Usually, beam-off events are the main source (up to ≈75%) for finding genetically correlated α particles originating from the decays of implanted nuclei with half-lives longer than ≈5 ms. The remaining ≈25% of α decays occur during the 5-ms beam-on periods. Events detected during beam-on periods without a coincident MWPC signal are also shown in Fig. 2. They were also taken into account in the correlation analysis.
In general, the analysis procedure to search for α-decay chains was the same as the one described in Ref. [16], except for the selected energy windows for the α-like events. It is noteworthy that the identification of α-decay chains originating from SHN with Z = 119 is supposedly relatively simple, since these SHN and their Ts daughters are expected to undergo α decay with half-lives less than 1 s [47] and decay properties of their progenies 287,288 Mc are known [1]. In the case of the Z = 120 nuclei, we do not exclude to observe α-decay chains starting only from the hitherto unknown daughters 291,292 Og (Z = 118), because the half-lives of the Z = 120 mothers could be very short [47,48]. This might lead to decay before the implantation in the stop detector has occurred. However, if these short-living Z = 120 nuclei can survive ≈0.6-μs flight time through TASCA, then the CANDI system would allow their decay to be resolved down to time differences of about 100 ns between the implantation and decay as was demonstrated for short-lived nuclei near the closed N = 126 shell region [41]. If the half-lives of Z = 120 nuclei were less than ≈0.6 μs then the calculated 55% TASCA transmission would be reduced by a factor of more than 2.
By taking into account various possible scenarios for the expected decay chains, we performed various positionand time-correlation analyses between implantation (ER), α, and fission (SF) events to find chains of nonrandom origin. Correlation analyses searching for ER- The energy conditions for the first and second α-like events were 8.5-13.0 MeV. As the energy ranges are the same and the searching times are long, the random correlation rate was relatively high, especially for α-like events detected during beam-on periods. However, these search conditions ensure that also all nonrandom decay chains with "missing" member(s), e.g., α particles escaping into the backward open hemisphere of the FPDS, will be found.
Any event with energies 3-20 MeV and coincident to an MWPC signal was considered to be an ER-like event [16]. Average counting rates of ER-like, α-like, and SF events per pixel of the stop detector during the beam-on and beam-off periods were similar to that for the 48 Ca + 249 Bk reaction (see Ref. [16]).
Only randomly correlated events similar to the ones found in the 48 Ca + 249 Bk reaction measured at TASCA were observed. A detailed discussion on the origin of random events is given in Ref. [16]. Finally, as a result of these analyses, no correlated ER, α, and SF events, having decay properties of SHN and originating from the expected α-decay chains of 295,296 119 and 295,296 120, were detected [1,16]. We determined a cross-section value for the observation of one event (hereafter: cross-section sensitivity) by taking into account the variations of the 249 Bk and 249 Cf target thicknesses over time as shown in Fig. 1, and the efficiencies of TASCA and the FPDS given in Sec. II. For the 50 Ti + 249 Bk reaction the crosssection sensitivity reached 65 fb, and for the 50 Ti + 249 Cf reaction it was 200 fb.
It is worth noting that these cross-section sensitivity levels are applicable only in the cases where the isotopes of elements 119 and 120 decay by α-particle emission. In the case of their fission decay for which recent calculations show non-negligible probabilities [49,50], the present experiment was not sensitive enough to identify an origin of fission events in the ER-SF correlation analysis. This was due to a large number of fission events (in total about 25 thousand) originating from the decay of targetlike nuclei produced in transfer reactions [16].

IV. DISCUSSION
In the present experiment relatively low cross-section sensitivities were reached, however elements 119 and 120 were not detected. The reached cross-section sensitivities together 064602-4    [1,[5][6][7][8][9][10][11][12][13][14][15][16]51] and with 197 Au [53]. Present experimental cross-section sensitivities for 50 Ti + 249 Bk and 50 Ti + 249 Cf are shown together with those of the 54 Cr + 248 Cm [23], 58 Fe + 244 Pu [22], 64 Ni + 238 U [21], and 48 Ca + 232 Th [52] reactions. Proton numbers of compound nuclei formed in these reactions are indicated. with the ones reached for three other reactions leading to element 120 [21][22][23] reveal the impact of the change of the projectile and of the compound nucleus's Z in the fusionevaporation reaction. Practically, it means a deviation of cross sections for reactions with heavier projectiles from the well-known 48 Ca-induced reactions. Accordingly, the results shall be discussed relative to the known properties of the 48 Ca-induced reactions. Such a comparative analysis was made in Refs. [18,19]. The maximum cross sections of fusionevaporation reactions with certain projectile or target nuclei exhibit well-pronounced trends as a function of the Coulomb parameter, Z p Z t /(A 1/3 p + A 1/3 t ). In Fig. 3(a), all known maxima of either σ 3n (max) or σ 4n (max) from the 48 Ca-induced reactions [1,[5][6][7][8][9][10][11][12][13][14][15][16]51] leading to the formation of SHN are shown as a function of the Coulomb parameter. Only crosssection sensitivity is known for the 48 Ca + 232 Th reaction [52]. In addition, the σ 3n value from the 48 Ca + 197 Au reaction [53] is shown to explore a systematic trend in a wider range of the Coulomb parameter. The present cross-section sensitivities together with those of the other three reactions leading to element 120 [21][22][23] are also shown in Fig. 3(a). It is important to mention that despite the different projectile and target nuclei all these reactions share one common feature: all target nuclei are deformed.
All features seen in Fig. 3(a) are due to properties of the fusion-evaporation reaction, which consists of three subsequent processes: capture, fusion, and the compound nucleus's deexcitation. Accordingly, the ER cross section is described by the three-term expression where σ cap is the cross section characterizing two captured nuclei forming a composite system at a collision energy E c.m. , P CN is the fraction of composite systems that forms a CN, and W CN is the survival probability of the CN against fission through particle evaporation at excitation energy E * = E c.m. − Q and angular momentum J. Q is the mass difference between the sum of reactant nuclei and the CN. In the case of the heaviest nuclei, mostly neutrons are evaporated [5,[54][55][56], leading to a preferential and almost exclusive population of neutron evaporation channels (xn channels). Thus, discussing ERs of the fusion reaction leading to SHE, we refer to all possible neutron-evaporation channels (xn channels). The terms σ cap and W CN describe independent processes and have been studied substantially, both experimentally and theoretically. Many theoretical calculations describe the known σ xn values of the 48 Ca-induced reactions fairly well. However, their predictions for the elements 119 and 120 have large deviations of several orders of magnitude. Thus, the predictive power of those calculations is limited and needs to be verified experimentally. Thus, the choice of any particular theoretical result as a baseline for the planning of a new-element search experiment becomes somewhat arbitrary. A systematic analysis of accumulated experimental data on the fusion-evaporation reactions as given in Fig. 3(a) could be useful for the planning of the experiment, and also for the discussion of the obtained results.
As seen from Eq. (1), essential for the observation of the desired ER is a proper choice of the beam energy (E c.m. or E * ). Our results, i.e., the nonobservation of ERs, do not allow any conclusion to be drawn about the proper choice of beam energy, which matched the maximum of a calculated 4n-evaporation excitation function of the 50 Ti + 249 Bk reaction [34]. The maximum cross-section values shown in Fig. 3(a) correspond to various projectile energies, each optimal for a given reaction and CN. In fact, in almost every 48 Ca +actinide reaction, ERs were observed at projectile energies up to 10 MeV above the fusion barrier (V B ) [57], i.e., (1.00-1.07)V B [1,58,59]. These energies then correspond to E * in the range of [35][36][37][38][39][40][41][42][43][44][45] MeV. This shows that E * within this range does not drastically affect the final results of these 064602-5 fusion-evaporation reactions, i.e., the σ ER . However, it does affect each term of Eq. (1). Projectile energies greater than V B ensure large σ cap , which drops drastically below the barrier and increases exponentially as a function of E c.m. [60]. The recently measured barrier distribution of the 48 Ca + 248 Cm reaction shows that the distribution's centroid is located at an energy slightly greater than V B [58]. This is due to the deformation of the target nucleus, which provides two distinct collision geometries: tip and side. The side collision, which is predicted to be the main source for fusion according to time-dependent Hartree-Fock calculations [61], results in a larger potential barrier for the fusion than the tip collision and V B calculated for spherically shaped nuclei [57]. Thus, observation of ERs in 48 Ca +actinide reactions at energies of (1.00-1.07)V B is seemingly also due to an increase in P CN as a function of E c.m. [62]. Finally, the survival probability of the CN is reduced with an increase of E * . Thus, overall, the three terms in Eq. (1) compensate each other as a function of E c.m. , which leads to a broad energy range for observing ERs.
From a nuclear reaction point of view (σ cap and P CN ), one can assume that the present 50 Ti + 249 Bk and 50 Ti + 249 Cf reactions are similar to the 48 Ca-induced reactions because of the target deformation. There should be some deviations due to the influence of the nuclear structure of the reactants, which largely impacts fusion-evaporation cross sections in Pb-target based reactions [60,[63][64][65] but is not yet fully understood in reactions with deformed target nuclei [66,67]. Nevertheless, broad ER excitation functions of the present reactions, similar to the ones for 48 Ca-induced reactions, can be assumed. Theory supports such a conclusion [34,[68][69][70][71][72]. The beam energy which we chose for the 50 Ti + 249 Bk and 50 Ti + 249 Cf reactions corresponds to ≈1.05V B (V B = 223.7 MeV) and 1.04V B (V B = 226.2 MeV), respectively. These values are within the above-mentioned energy range where fusion is predicted to be enhanced according to the results from the experiment [58] and the theory [61]. Therefore, it is unlikely that the used beam energy was the major reason for the nonobservation of elements 119 and 120. More likely, it is due to the very low cross sections of these reactions, which have their origin in the fusion-reaction mechanism and the survival probability of the fused system.
The fusion probability of the two nuclei predominantly depends on the Coulomb force between the reactants. The composite system, which consists of many protons and neutrons, often fails to form a CN under the influence of the resulting total Coulomb force and the system reseparates. This so-called quasifission (QF) process has been known for decades and has been investigated in detail [17,25,66,[76][77][78][79]. The presence of QF (P CN < 1) had been predicted theoretically for reactions with a projectile-target charge product Z p Z t 1600 [76]. To date, this limit has been altered and QF has been proven to also be present in reactions having much smaller Z p Z t values [63,[80][81][82]. All reactions given in Fig. 3(a) have Z p Z t values exceeding the original threshold value (except for the case of 197 Au). Thus, QF can be expected to be dominant over fusion as was also experimentally shown [17,66,78,79,83,84]. In general, a quantitative estimate of the QF probability as a function of Z p Z t is a very complex problem and intensive research on this topic is still ongoing.
In Refs. [34,83], semiempirical estimates are given that are based on the P CN values deduced from the experimental data.
By using the expression given in Ref. [83] to estimate the P CN values for actinide-target based reactions, P CN was calculated for all reactions shown in Fig. 3(a). The results are shown in Fig. 3(b). P CN values are exponentially decreasing as a function of the Coulomb parameter. This is in agreement with the trend of the σ xn values, which decreases exponentially [18,19] until Hs [cf. Figs. 3(a) and 3(b)]. It also shows that a change of a few protons in the target's Z with a fixed projectile nucleus does not lead to a significant variation of the Coulomb parameter because of the large numbers of initial protons. For instance, a change from 232 Th (leading to formation of 280 Ds * ) to 243 Am ( 291 Mc * ) is a 4.7% change in the target's Z, which results in a decrease in P CN by a factor of about 3 in the 48 Ca-induced reactions.
However, in the 48 Ca + 232 Th reaction, for which fusion is more favorable, the element Ds was not observed at a cross-section sensitivity of about 9 pb [52], which is close to the maximum σ xn value obtained for the 48 Ca + 243 Am reaction [1,13]. In fact, for the reactions leading to SHEs heavier than Ds the trend of an exponential decrease in σ xn values is broken and the maximum cross section remains at a level of ≈0.5-10 pb with a local maximum at Z ≈ 114. This feature, i.e., a sudden increase in the ER cross sections relative to a decreasing trend, was discussed in connection with the survival probability of the CN [1,18,19].
Once the compact-shaped CN is formed and at full equilibrium in all internal degrees of freedom as a result of fusion, it has an excitation energy that will be released via fission, emission of light particles, and/or electromagnetic transitions [57,62]. Each reaction shown in Fig. 3 has been measured at CN excitation energies greater than 30 MeV, which significantly exceeds the height of the fission barrier, B f , and the neutron and the proton separation energies of the CN. The survival probability of the superheavy CN is often expressed as where T and x are the temperature and the number of emitted neutrons, respectively. Here, one should note that the successive emission of neutrons at each ith step forms a "new" CN, which has to survive against multichance fission processes. Recently, the interest in multichance fission in the deexcitation of the CN has been renewed due to experimental evidence showing significant contributions of this process in fusionfission [63] and transfer-induced fission [85,86] reactions. However, the crucial point remains the survival probability of the initially formed CN, where both E * and J are high [63,85]. Accordingly, the B f of the initial CN is one of the important quantities for the survival process [18,19,63,85].
To shed some light on the dependence of the σ xn value (which reflects the survival probability of the excited CN) on the CN's fissility, we show in Figs. 3(c) and 3(d) calculated B f values for the CN formed in each reaction shown in Fig. 3(a). We selected four different theoretical values for the ground-state fission barriers as representatives of two main 064602-6 theoretical approaches, i.e., macro-microscopic [48,73] and microscopic [74,75] ones. Presently, many other theoretical predictions performed within these two approaches exist (see Refs. [87][88][89][90] and references therein) and their results are equally valid as those selected in this paper and will not affect the quality of the discussion.
All calculations predict an increase of B f in nuclei of the elements above Ds. This indicates that their stability against fission is enhanced due to the shell effects related to the island of stability [1,18,19]. Such an increase in the predicted B f values can be the reason for a sudden increase in the measured ER cross sections, which are due to the enhanced survival probability. Moreover, the nonobservation of Ds nuclei in the 48 Ca + 232 Th reaction at the cross-section level of about 9 pb can be explained by their low B f value, which is barely affected by the shell effects originating from the island of stability.
However, the results of different theoretical frameworks deviate in the prediction of the B f values for SHN, and at which Z the next shell closure will occur. According to macromicroscopic models, the effect of the enhanced fission barrier is most dominant in Fl nuclei and decreases towards heavier SHN. A similar trend is observed in experimental σ xn values as shown in Fig. 3(a). The self-consistent purely microscopic models predict the B f values for the elements 119 and 120 to be similar to or greater than those in the range of Fl to Og. Overall, the results obtained in both frameworks are able to explain the observed trend of the σ xn values up to Og. Supposedly, an increase in B f values of these SHN results in an increase of W CN , which counteracts a reduction in P CN . In this case, the above-mentioned factor of about 3 in decrease in P CN from 48 Ca + 232 Th to 48 Ca + 243 Am may be compensated by an increase in the B f values in both models. However, not in all calculations B f values significantly increase beyond Og. This leads to the assumption that the upper limit for the gain in W CN for SHN with Z = 119 and 120 is the same as for the 48 Ca-induced reactions. However, the question is whether this gain can completely compensate a reduction in P CN or not.
The present two reactions show a reduction in P CN by a factor of more than 5 compared to any of the measured 48 Ca-induced reactions. Such a strong reduction is due to the relatively large change in the Coulomb parameter that originates from the ≈4.2 % change in the projectile's Z. One can estimate that the reduction in σ xn is at least about a factor of ≈5.6, which is the relative decrease in P CN for 50 Ti + 249 Bk compared to the 48 Ca + 249 Cf reaction, where the element Og was produced with σ 3n ≈ 0.5 pb. This translates to the necessity to reach a cross-section sensitivity on the order of 90 fb. This value indeed has been reached in this experiment for element 119 (65 fb). However, the absence of any detected event indicates that σ xn may decrease stronger than suggested by the P CN estimation given in Ref. [83]. A rapid decrease in P CN would favor the results of the microscopically calculated B f shown in Fig. 3(d) for the synthesis of the elements beyond Og. At the same time, the macro-microscopic results for B f do not exclude a further reduction of σ xn for the syntheses of the elements 119 and 120 in each projectiletarget combination in addition to the reduction due to the change in P CN . Eventually, this raises interest in measuring cross sections of the 48 Ca + 254 Es and 48 Ca + 257 Fm reactions. This, however, is presently impossible [20]. According to the macro-microscopic calculations one may expect a strong reduction in σ xn values. At the same time according to microscopic calculations one could expect to observe reasonably high σ xn values.
Overall, it is evident that the projectile-target combination primarily defines the fusion probability. As seen in Fig. 3(b), the fusion probabilities of the other three reactions leading to element 120 will be further reduced. One can estimate that the reaction 54 Cr + 248 Cm will probably have ≈3.6 times smaller σ xn values than the 50 Ti-induced reactions, which will require reaching a cross-section sensitivity below 20 fb. Consequently, the reactions with heavier projectiles ( 58 Fe, 64 Ni, etc.) will suffer from a further reduction of P CN . Therefore, 50 Ti-induced reactions are the most promising for exploring the discoveries of elements beyond Og. However, a final justification of these predictions has to come from the observation of the elements 119 and 120.
For the synthesis of element 119 another potential reaction is 51 V + 248 Cm, which leads to the same CN as the 50 Ti + 249 Bk reaction. According to the discussions above, one could expect a roughly two times smaller P CN for the 51 V + 248 Cm reaction compared with 50 Ti + 249 Bk for which presently reached cross-section sensitivity is 65 pb. This would require one to perform an experiment for the former reaction at the cross-section sensitivity on the order of at least 30 fb at a beam energy corresponding to the maximum of either the 3n or the 4n channel.

V. SUMMARY AND CONCLUSION
A four-month long experiment with a high-intensity 50 Ti beam bombarding 249 Bk/ 249 Cf targets was carried out successfully at the gas-filled recoil separator TASCA. In the data collected during this challenging experimental campaign, we searched for correlated α-decay chains from isotopes of the new elements 119 and 120. We did not observe any nonrandom α-decay chain terminating with a fission that could be attributed to the decay of superheavy nuclei with Z = 119 and 120. This resulted in "one event" cross-section sensitivities of 65 and 200 fb for the 50 Ti + 249 Bk and 50 Ti + 249 Cf reactions, respectively, at a midtarget beam energy of E lab = 281.5 MeV.
These cross-section sensitivity levels are not applicable if the isotopes of the elements 119 and 120 would directly decay either by spontaneous fission and/or electron-capture delayed fission for which recent calculations show non-negligible probabilities [49,50]. In the present paper unambiguous identification of such direct fission decays, i.e., ER-SF from SHN, was not possible. However, in the future one should consider the search for unknown spontaneous fission or electroncapture delayed fission branches of SHN, which would reduce the identification efficiency of SHN via α-decay chains.
The nonobservation of isotopes of elements 119 and 120 was discussed within the context of the fusion-evaporation reaction mechanism. It is apparent that the fusion probability of 50 Ti and heavier projectiles with actinide targets is significantly reduced compared to 48 Ca-induced reactions. The present combinations, 50 Ti + 249 Bk and 50 Ti + 249 Cf, are still 064602-7 considered as the most promising reactions for the syntheses of the next two elements beyond Og. However, for the observation of evaporation residues from these reactions, experiments need to be carried out that reach much smaller cross-section levels, as compared with the ones presently obtained.

ACKNOWLEDGMENTS
We are grateful for GSI's Penning ion source and UNI-LAC staff, and the Experimental Electronics department for their continuous support of the experiment. This work was financially supported in part by the German BMBF (Grant