Discovery of the elements with atomic numbers Z = 113, 115 and 117 (IUPAC Technical Report)

3.1 Z=113

A confounding feature of alpha decay in odd-Z nuclei is the likely broad distributions of α-energies due to decays to excited daughter states. A good illustration can be found in those of ²⁷²Rg studied by Morita et al. [7]. The results indicate a distribution of α-energies as much as 0.8 MeV. The clustered transitions show nearly the same level lifetime suggesting that each cluster of α-decay transitions originates from a common parent state. This is consistent with the view that the different α-energies, negating reproducibility when few events are available, correspond to split daughter states of odd–odd or odd-A nuclei.

The collaboration of Oganessian et al. [8] in 2004 studied the hot fusion reaction of ⁴⁸Ca with ²⁴³Am and reported three chains beginning with ²⁸⁸115, continuing sequentially to ²⁸⁴113, ²⁸⁰Rg, ²⁷⁶Mt, ²⁷²Bh, were observed with good internal agreement and terminating with spontaneous fission of ²⁶⁸Db with mean lifetime of 23 h. At a different bombardment energy, one four member α-decay chain commencing at ²⁸⁷115, passing through ²⁸³113 and then ²⁷⁹Rg and ²⁷⁵Mt leading to ²⁷¹Bh whose missing α-decay was inferred to lead to ²⁶⁷Db which decayed by spontaneous fission (with a 100 min lifetime). Data are summarized in the Table 1. None of the nuclides had been previously characterized.

The production of two chains of α-emitting nuclides was reported by Morita et al. from the cold fusion reaction of a bismuth-209 target with a ⁷⁰Zn beam at the RIKEN heavy-ion facility in Japan, the first in 2004 [9] and the second in 2007 [10]. The former study reports the α-chain commencing with ²⁷⁸113 proceeding through ²⁷⁴Rg, ²⁷⁰Mt, ²⁶⁶Bh, and terminating via spontaneous fission decay assigned to ²⁶²Db. All α-energies and lifetimes were measured. In the subsequent study, a very similar sequence was found but with some reproducibility difficulties. The full α-energy for ²⁷⁰Mt was not measured; those for ²⁶⁶Bh were in disagreement (9.08 vs. 9.77 MeV); and the lifetimes for ²⁶²Db spontaneous fission were suggestively different (41 s vs. 0.8 s). For both chains, position-sensitive detectors were used as is the case in all investigations reviewed here. These provide a high degree of confidence that the observed decays are indeed sequential decays in each case. Nuclides reported in these chains do not correspond to established (recognized) isotopes (see the following figure). But a report of a single triple-coincidence of α-emitters commencing with the useful intermediate ²⁶⁶Bh had been described by Wilk et al. [11]. Production was via the hot fusion ²²Ne+²⁴⁹Bk reaction, and the leading event had an α-particle energy of 9.29 MeV, within the RIKEN results, with an estimated half-life of 1–10 s. It was followed by a 28 s α-decay, not by spontaneous fission. The latter observation is in contrast to the RIKEN result.

PRIOR JWP ASSESSMENT: The work of the collaboration of Morita et al. [9, 10] is very promising but has not met the Criteria for discovery owing to the paucity of events, the absence of firm connection(s) to known nuclides, and the inconsistencies noted above.

An investigation by Qin et al. [12] in 2006 at the Heavy Ion Research Facility in Lanzho, People’s Republic of China used the ²⁴³Am(²⁶Mg,3n)²⁶⁶Bh reaction to produce four chains with a 1 s lifetime and ²⁶⁶Bh alpha energy (9.03±0.08) MeV. These were followed by (8.53±0.07) MeV alpha decays to a 37.5 s lifetime ²⁶²Db with one of the four chains continuing to ²⁵⁸Lr in 5.07 s with a 8.64 MeV alpha decay.

In 2007, the collaboration of Oganessian et al. [13] investigated the hot fusion of ⁴⁸Ca with ²³⁷Np and reported two four-member α-decay chains commencing at ²⁸²113, passing through ²⁷⁸Rg, ²⁷⁴Mt, and ²⁷⁰Bh, and leading, in just one chain, to ²⁶⁶Db decay by spontaneous fission with a 32 min lifetime. The first two events in each chain showed excellent mutual agreement for both decay energies and lifetimes.^1[1] The third member gave lifetimes of 470 and 810 ms. None of the nuclides had been previously characterized.

PRIOR JWP ASSESSMENT: The 2004 collaboration of Oganessian et al. at Dubna was approximately contemporaneous with that of Morita et al. at RIKEN. The Dubna results in combination with the 2007 collaboration are encouraging but do not meet the Criteria for discovery because of the paucity of events, the lack of connections to known nuclides, and the absence of cross-reactions.

In 2009, Morita et al. used the fusion reaction ²³Na+²⁴⁸Cm to produce 16 persuasive chains of ²⁶⁶Bh providing a cross reaction to that alpha-emitting nuclide and its descendants in the Z=113 alpha-decay chain [14]. The average alpha energy from ²⁶⁶Bh was 9.12 MeV. For seven of those, alpha decay to ²⁶²Db had an average 8.67 MeV energy and for four cases the chains continued to the 4.1 s alpha decay of ²⁵⁸Lr [15] with an average energy of 8.67 MeV all with excellent internal agreement and with the 2006 study of Qin et al. [12]. These agree with the previously determined property [15] of 4.1 s ²⁵⁸Lr decay (4 alpha branches between 8.59 and 8.68 MeV) and serves well to establish the ²⁶⁶Bh as a known anchor.

In 2012, the Morita et al. collaboration [16] repeated their earlier experiments and observed a third decay chain reportedly commencing with ²⁷⁸113. The properties of all three chains are reproduced in Fig. 2 below derived from [16].

Fig. 2:

Summary of the Z=113 decay chains observed by the Morita et al. collaborations (q.v.). Values are E_α/MeV above lifetimes in milliseconds (ms) or seconds (s).

In the case of Morita’s Z=113 chains, (the first two previously assessed as “inconsistent”) the sum of the α₁ and α₂ energies is nearly the same, sidestepping the otherwise troublesome individual differences. Furthermore, the sum of α₁+α₂+α₃+α₄ energies is also nearly the same for the first and third chains,

signaling that the chains reasonably start and end up at the same states in the same nuclides. This appears to indicate that the starting and finishing states in the transition from ²⁷⁸113 to ²⁶²Db are the same in the two chains, even if different states are involved in the decays in between. This approach, although not decisive, can be applied to many odd-nuclei examples and its success is a sufficient, but not necessary condition for supporting consistency where it might be concealed by decays through a choice of intermediate states particularly when few events are recorded. We note that the total decay chain energies are not guaranteed to agree as in some cases gamma ray and/or conversion-electron energies may not be recorded and that these totals, although not imperative, can prove useful.

An important consideration is if the observed chains of Morita et al. are convincingly anchored to known nuclides. This is demonstrated in the Table 2 confirming that the 2012 decays α₅ and α₆ are in concordance with other determinations: the ²⁶⁶Bh study by Morita et al. using the ²³Na+²⁴⁸Cm fusion reaction and the decay of ²⁶²Db summarized in the 2001 Nuclear Data Sheets Akovali [15].

Conclusion: The observed Z=113 A=278 chain decay energies for ²⁶²Db (α₅) and ²⁵⁸Lr (α₆) are consistent with those measured in the Morita et al. (2009) ²⁶⁶Bh study and with values observed earlier by others [15]. Also, the lifetime values (not shown here) given in Morita et al. (2012) were consistent with the lifetimes measured previously [15]. All the “inconsistencies” pointed out in the previous JWP ASSESSMENT except for the missing α₃ energy in the second chain have been resolved.

JWP ASSESSMENT: Three chains of ²⁷⁸113 observed by the RIKEN collaborations, the first in 2004 [9], the second in 2007 [10], and the third in 2012 [16], are now construed as being consistent. Firm connection to established nuclides is provided. The remaining criterion achieved for acknowledgement of discovery is an identification of Z which is now embodied in the cross reaction production and characterization of the chain beginning with ²⁶⁶Bh as found by the RIKEN collaboration in 2009 [14] and by Qin et al. in 2006 [12]. The Criteria for discovery have been met.

The 2010 collaboration of Oganessian et al. [17–19] used hot fusion ²⁴⁹Bk(⁴⁸Ca, xn) reactions to produce 5 alpha-decay chains originating with ²⁹³117 passing through ²⁸⁵113 and one additional chain from ²⁹⁴117 passing through ²⁸⁶113. In 2012 [20] and 2013 [21], Oganessian et al. used the ²⁴³Am(⁴⁸Ca,2n) reaction to observe ²⁸⁹115→²⁸⁵113, a cross reaction verification of identifying ²⁸⁵113. But because the earlier collaboration discovery claims for ^282,283,284113 are different from the newly identified ^285,286113, these do not serve to confirm the earlier priority claims.

The 2013 collaboration of Oganessian et al. [21] used the hot fusion ²⁴³Am(⁴⁸Ca, xn) reactions to produce 24 persuasive alpha-decay chains originating with ²⁸⁸115 passing through ²⁸⁴113, 4 alpha-decay chains originating with ²⁸⁹115 passing through ²⁸⁵113, and one incomplete chain from ²⁸⁷115 passing through ²⁸³113. Complementary to this, the 2013 collaboration of Rudolph et al. [22] used the same reaction at GSI (Darmstadt) to produce 6 complete chain sequences in excellent agreement with the Dubna Oganessian collaborations. Despite the relatively large productivity, the energy dependence providing a picture of the excitation function is not illuminating. That is, “a yield curve: production cross section as a function of energy of the particle impinging upon a target nucleus” [1] was not considered statistically compelling.

The 2013 Oganessian collaboration [21] and the 2013 Rudolph collaboration provide redundancy to the three ²⁸⁴113 chains observed in 2004 with the alpha energies being in excellent agreement among most of the events. Table 3 and 4 summarizes alpha energies in MeV for the 24 and six events in the 2013 experiments and the three events from the earlier experiment.

Much of the minor discrepancies in energy are accommodated when sums are considered.

However, the Criteria (q.v. [1]) have not been met as there is no mandatory identification of the chain atomic numbers neither through a known descendant nor by cross reaction. Chemical determinations as detailed in the subsequent profile of Z=115 where they are documented, serving the important role of assigning atomic number are insufficiently selective although certainly otherwise informative.

JWP ASSESSMENT: The 2004 collaborations of Oganessian et al. at Dubna and those of Morita et al. at RIKEN were published on 2 February and 15 October, respectively. Recent results from the RIKEN group have cross reactions that support anchoring their claims to known atomic numbers. For Dubna, many alpha-decay chains from the ²⁴³Am(⁴⁸Ca,3n)²⁸⁸115 fusion serve to duplicate and reinforce the characteristics assigned to ²⁸⁴113 observed earlier. But, the 2004 and 2007 Dubna physical measurements were not able to within reasonable doubt determine Z. Chemistry probes that had been employed in previous experiments remain unpersuasive in their ability to assure the Z value. (Details are discussed below in the consideration for Z=115.) No new chemistry separations or simulations have been instituted to nullify that JWP criticism although preliminary studies have been done to explore alternatives [23]. We conclude the Criteria for discovery have not been met by the Dubna-Livermore group claims for Z=113 while the RIKEN claim does meet the Criteria. Priority for the discovery has been fulfilled by the RIKEN collaboration.

3.2 Z=115

The 2004 collaboration of Oganessian et al. [8] which was mentioned in the Z=113 discussion above reports one chain commencing with ²⁸⁷115. In this same fusion experiment, at a slightly lower beam energy and the same beam dose, three new consecutive α-decay chains were reported assigned to the ²⁸⁸115 isotope and products ²⁸⁴113, ²⁸⁰Rg, ²⁷⁶Mt, and ²⁷²Bh with agreement among the five sets of α-particle energies and among the five lifetime values. All terminated in a approximately 26 h spontaneous fission lifetime assigned to ²⁶⁸Db. All five nuclides were reported for the first time.

The collaboration of Dmitriev et al. [24, 25] at the Flerov Laboratory in Dubna used the ⁴⁸Ca+²⁴³Am fusion reaction in 2005 to produce 15 additional spontaneous fission nuclides and other reaction products recoiling directly, without selectivity, onto a copper surface from which they were extracted and subjected to cation exchange separation. The technique employed was claimed to distinguish between Group 3 elements (lanthanides and actinides) and combined Groups 4 and 5. Theoretical expectations were proposed for discounting Group 4, corresponding to Rf spontaneous fission. Yet, theoretical predictions were also noted for the possible inversion of the sequence of trends in a periodic group amongst the heavy elements. The mean lifetime for the assigned spontaneous fission averaged 46 h for the 15 nuclides, a value within statistical agreement with the first determinations by Oganessian et al. [8].

In 2007, the Livermore-Dubna collaboration of Stoyer et al. [26] published their production of ²⁸⁸115 from the ⁴⁸Ca+²⁴³Am fusion reaction recoiling directly onto a copper surface and to extract the terminal, long-lived ²⁶⁸Db for revised chemical separation procedures. Two procedures were employed, each aimed at separating Group 4 from Group 5. A total of five spontaneous fission events were observed with lifetimes of 16–37 h. In those separations that were also able to distinguish between Nb and Ta-like chemical behavior, all three such events appeared in the Ta-like fractions.

PRIOR JWP ASSESSMENT: The Dubna–Livermore collaborations have reported a total of 23 events assigned either directly or indirectly to ²⁸⁸115 via a single target-projectile combination. The assignment is supported mostly by chemical studies of the terminal spontaneous fission assigned to ²⁶⁸Db. Those chemical studies serve a central role in whether or not the Criteria of identification have been met. No carrier-free actinide tracers were employed despite the extremely complex oxidation chemistry and adsorption quirks of those Group 3 elements in contrast to lanthanide behavior. Chemical properties of confirmed heavy elements are important for improving relativistic theories of chemical behavior. However, by itself, current theory is sufficiently uncertain that it cannot be used to distinguish the properties of Groups 4 and 5 elements in this region with confidence.

In 2007, one-and-a-half years after the Dubna–Livermore [26] experiments, the chemically separated samples assigned in 2005 as dubnium were measured for very long periods of time in an alpha spectrometer by a group at Livermore [27]. The alpha spectra revealed the presence of Am, Cm and Po isotopes, the latter presumably produced by transfer reactions on Pb impurities in the target or target backing. A Dubna study of one 2-year-old counting sample by Dressler et al. [28] found alpha activity due to some americium isotopes but no evidence of spontaneous fission activity during very long counting times. Despite separations designed to occlude interferences, the extraordinarily low event rates and as yet not perfected chemistries counsel the need for extreme caution in the face of uncertain chemical purity.

The 2007 collaboration of Oganessian et al. [13] reported the observation of two decay chains originating from ²⁸²113 produced in the hot fusion ²³⁷Np(⁴⁸Ca,3n) reaction. This does not overlap with the ²⁸⁷115 and ²⁸⁸115 originating Z=113 decays observed earlier from hot fusion ²⁴³Am(⁴⁸Ca,4n) and ²⁴³Am (⁴⁸Ca,3n) reactions, respectively.

The 2012–2013 collaborations of Oganessian et al. [20, 21] produced a large number of five-membered chains again with the ⁴⁸Ca+²⁴³Am combination. Twenty-four chains beginning with ²⁸⁸115 decayed with alpha-particle emission of average energy (10.46±0.06) MeV followed by the descendants listed in the previous table that agree with excellent precision with the 2004 results. In this collaboration [20, 21], four decay chains originating from ²⁸⁹115 were produced in the (⁴⁸Ca, 2n) reaction channel. These results are summarized by Oganessian et al. [21] and in Figure 3 below.

Fig. 3:

Complete decay chains observed by Oganessian et al. derived and edited from figure 5 of that reference [21]. The number of events (publication dates) in bold at the bottom corresponds to those recognized by the JWP as persuasive. For each isotope E_α/MeV and lifetimes in milliseconds (ms), seconds (s) or hours (h) are shown.

CHEMISTRY: An essential issue for the JWP’s consideration was that if the chemical behavior could be unequivocally assigned to Z=115, that would have been a profound decider on the identity of its precursor chain members. A number of chemistry experiments were performed on the long-lived spontaneously fission product recovered from Ca+Am fusion reactions. Its half-life is approximately 26 h. However, the chemical properties of (Rf and) Db can be significantly different from their corresponding lighter congeners, Zr-Hf and Nb-Ta, respectively. Thus, the traditional simple congener principle may not be applied with robust confidence to identify Db. Use of relativistic modeling of chemical behavior to circumvent such departures is an extraordinary new venture and needs definitive observed behavior to bolster its applicability, but that theory-experiment convergence is still in the future for high-Z elements despite success at lower atomic numbers.

Starting in 1988, a number of chemical studies of dubnium were performed in which about 30 s half-life ^262,263Db were both unquestionably produced by the ²⁴⁹Bk(¹⁸O,xn) reactions in which x=4 and 5. These investigations looked at the extraction of Db in strong acid organic ligand partitions, the latter being methyl isobutyl ketone [29], diisobutylcarbinol [30], triisooctylamine [31], Aliquat 336 [32], and in cation exchange using α-hydroxyisobutyric acid [33]. As in all of the published separation schemes, radiotracers are added to follow congeners niobium and tantalum and lanthanide radiotracers serve as proxies for actinide behavior. But the lanthanides chosen are restricted to those with 3+ oxidation states yet actinides and some lanthanides are commonly found in 4+ or 5+ states. Furthermore, the actinide protactinium has sometimes, but not always, been used as a quasi-homologue of a Group 5 element. In the separation studies, dubnium at times behaves more like protactinium than like either niobium or tantalum. What is seen is that even in the ¹⁸O-irradiation investigations where Db is definitely identifiable the possibility of interferences has not been eliminated. A 2005 study by Schumann et al. [34] employed a wide range of tracers including from Groups 6 and above to serve as homologues of heavier transactinide congeners (despite recognition that relativistic effects can frustrate such expectations). One of the observations of that study was that a single initial precipitation step in the dubnium separations was contaminated by a variety of elements expected to have been occluded. Wilk et al. [35] sought to improve the separation scheme employed by Stoyer et al. [26] and encountered Nb and Ta both evincing two behaviors which could not be explained although the authors did suspect that this was evidence for different speciation of both the Nb and Ta in the solutions employed. The existence of a variety of halide, oxyhalide, and hydroxyhalide forms for the heavy Group 5 elements has been acknowledged as a complicating factor in their chemistries [36–38]. The chemistry employed in the Z=115 studies seeking to prove the identity of dubnium did not employ the same steps as in the definitive studies on the short-lived ²⁶²Db and ²⁶³Db.

The Oganessian collaborations of 2010 studied the ⁴⁸Ca+²⁴⁹Bk fusion reactions to form Z=117 products [17, 18]. Three of the chains from the 4n evaporation channel included descendants ²⁸⁹115 and ²⁸⁵113. One transitioned through ²⁹⁰115 and ²⁸⁶113. The 2012 results of Oganessian et al. [20] for ⁴⁸Ca+²⁴³Am included one chain through ²⁸⁹115 and ²⁸⁵113 from the 2n evaporation channel and subsequently a total of four inclusive chains from the 2n evaporation channel were found in 2013 [21]. The average alpha-particle energies observed are presented in Table 5. Consequently, cross reactions show assigned decay characteristics of ²⁸⁹115(α)→²⁸⁵113(α)→²⁸¹Rg(SF) observed in the ²⁴³Am(⁴⁸Ca,2n) reaction (4 events) agree well with the assigned decays originating from ²⁹³117 observed in the ²⁴⁹Bk(⁴⁸Ca,4n) reactions (6 events).

Table 5:

Comparison of alpha-particle decay chains with E_α/MeV and half-lives in seconds (s) or hours (h) beginning with ²⁹³117 and ²⁹⁴117.

The energy sums for the entire six-membered alpha chains above are 57.74, 57.90 and 57.85 MeV, respectively.

The 2013 GSI collaboration of Rudolph et al. [22] studied the ²⁴³Am+⁴⁸Ca reactions. They reported the observation of 22 more decay chains originating from ²⁸⁸115 produced by the 3n evaporation channel terminating in the spontaneous fission of ²⁶⁸Db and one of ²⁸⁷115 by the 4n evaporation channel. Those reinforce the results of the Russia–US collaboration [21]. The same collaboration attempted to measure coincident X-rays whose assignment property would have been a significant boost to confidence in Z identification which is lacking in the Am+Ca production channel. The use of X-rays for elemental identification is complicated by the fact that the observed photons, expected to have very precise X-ray energies, might be γ-rays or Compton scattering or background events. Given the low number of photons detected at the predicted X-ray transition energies and the presence of other X-rays or γ-rays in the spectrum, the X-ray identification was not regarded as persuasive. The paper was able to demonstrate the coincidence between alpha decay and gamma-ray emission in the decays of ²⁸⁰Rg and ²⁷⁶Mt, which confirms the population of excited states in alpha decay of odd-Z super heavy nuclides.

JWP ASSESSMENT: The 2010 [17, 18] jointly with the 2013 [21] collaborations of Oganessian et al. have met the Criteria for discovery of the element with atomic number Z=115 in as much as the reproducibility of alpha chain energies and lifetimes of ²⁸⁹115 in a cross reaction comparison is very convincing.

3.3 Z=117

In 2010, the Russia-US collaboration of Oganessian et al. [17, 18] used the ⁴⁸Ca+²⁴⁹Bk fusion reaction to produce the compound nucleus ²⁹⁷117 leading to three complete four-member chains commencing with ²⁹³117 and one seven-member chain commencing with ²⁹⁴117 the last member of each ending with spontaneous fission. In 2012 a second series of ⁴⁸Ca+²⁴⁹Bk fusions to form the compound system ²⁹⁷117 were studied [19] resulting in three more four-member chains commencing with ²⁹³117 and two more seven-member chains commencing with ²⁹⁴117. Four more complete chains were measured by Oganessian et al. in 2013 including a new alpha decay from two of the events containing ²⁸¹Rg [39]. As noted in discussion of Z=115, the ⁴⁸Ca+²⁴³Am fusion reported by Oganessian et al. to have produced four short chains of ²⁸⁹115 [21] agrees with the daughter events from ²⁹³117 decay within the precision of the alpha-particle energies [17–19, 39]. That cross reaction comparison is included in the table below.

JWP ASSESSMENT: A convincing case in cross reaction producing ²⁸⁹115 and ²⁸⁵113 from both ⁴⁸Ca+²⁴⁹Bk and ⁴⁸Ca+²⁴³Am is demonstrated in the top of the previous table. Thus, the 2010 [17, 18], 2012 [19] and 2013 [39] jointly with 2013 [21] collaborations of Oganessian et al. have met the Criteria for discovery of the elements with atomic numbers Z=115 and Z=117.