Significant progress in the synthesis of the heaviest man-made chemical elements was achieved in the last decades resulting in the completion of the 7th row in the Periodic Table of the Elements (PTE) (Oganessian and Utyonkov, 2015). Due to decreasing nuclear fusion- and survival probabilities with increasing atomic number Z, the production rates of the superheavy elements (SHE) decrease rapidly, reaching a level of a single atom per month for oganesson (Og, element 118) (Oganessian and Utyonkov, 2015). The commonly used technique for SHE isolation is the kinematic separation in a recoil separator. The time between the production of a nuclear fusion reaction product in the target and its implantation as a recoil ion in the detection system is about 1 microsecond, allowing studies of nuclear decay properties of very short-lived isotopes. Experimental methods for chemical studies of SHE are more demanding and time-consuming, and therefore less efficient, especially for short-lived SHE isotopes. This makes studies of chemical properties more challenging (Türler and Pershina, 2013). Among the chemical techniques applied for SHE, a fast and efficient chemical separation can be achieved by gas-solid chromatography studies of volatile species (Türler and Pershina, 2013; Türler et al., 2015).

Chemical properties of yet unstudied SHE can be extrapolated based on the general law of periodicity, which connects the electronic structure of the elements with their position in the PTE. However, with increasing nuclear charge of the SHE, the influence of increasingly important relativistic effects on chemical properties has to be taken into account (Schwerdtfeger et al., 2020; Pyper, 2020; Pershina, 2019; Pershina, 2015). Apparently, the strong contraction of the s and p_1/2 atomic orbitals (AO) and a large spin-orbit (SO) splitting of the electron shells with the orbital angular momentum quantum number l ≥ 1, can lead to properties different from those of their lighter homologues. Thus, reliable predictions of chemical properties of SHE become of extreme importance for the heaviest members of groups 12–18 (Schwerdtfeger et al., 2020; Pershina, 2019; Smits et al., 2024). Since the reactivity of atoms is generally connected to the energy level of their valence AO, relativistic calculations of the AO energy levels and of the first ionization potentials (IP) allow assessing the validity of trends within the groups of the PTE (Eliav et al., 2015; Borschevsky et al., 2015; Dzuba and Flambaum, 2016; Trombach et al., 2019). A strong relativistic stabilization of the spherical atomic orbitals with closed-shell 7s² and quasi closed-shell 7p_1/2 ² electronic configurations, together with the large spin-orbit splitting within the 7p shell were shown to result in low AO energies and large IP values for Cn and Fl, respectively (Schwerdtfeger et al., 2020; Pershina, 2015; Eliav et al., 2015). Accordingly, these elements were predicted to be the least reactive members of groups 12 and 14 (Pitzer, 1975; Schwerdtfeger and Seth, 2002). These predictions were recently confirmed experimentally (Eichler et al., 2007; Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022).

The neighbouring odd-Z elements nihonium (Nh, element 113) and moscovium (Mc, element 115) are presently in the focus of theoretical studies. Since Nh and Mc possess one unpaired electron in the 7p_1/2 or 7p_3/2 subshells, respectively, higher AO energies and lower IP values than for Cn and Fl are predicted for Nh and Mc. This implies that the reactivity of Nh and Mc is higher compared with that of their closed-shell neighbours Cn and Fl. Therefore, chemical reactivity, i.e., the tendency to form chemical bonds, is predicted to be higher for Nh and Mc than for Cn and Fl (Schwerdtfeger et al., 2020; Pershina, 2015; Trombach et al., 2019). However, experimental data on the reactivity of Nh and Mc are required to validate the predicted local minimum at Fl. Indeed, Nh and Mc are expected to form chemical compounds, e.g., hydroxides, which should adsorb more strongly on surfaces than pure atoms (Pershina et al., 2009; Pershina and Iliaš, 2019). Recently, motivated by experimental developments (Yakushev et al., 2021), the interaction of Nh and Mc with Au and quartz surfaces was studied on the basis of periodic relativistic density functional theory (DFT) calculations (Pershina, 2016; Pershina et al., 2021). As a result, equal values of the adsorption energy/enthalpy (E _ads = −ΔH _ads) on quartz, of 58 kJ/mol, were predicted for elemental Nh and Mc (Pershina et al., 2021). Significantly (by about 100 kJ/mol) higher E _ads values were predicted for the adsorption of these elements on the Au (111) surface in several theoretical works (Trombach et al., 2019; Pershina, 2018; Rusakov et al., 2013; Fox-Beyer and van Wüllen, 2012). The obtained E _ads values for Nh and Mc turned out to be significantly (by about 50–100 kJ/mol) lower than those of their lighter homologues, Tl and Bi, respectively, on both types of the surfaces, quartz and Au. This is due to the strong relativistic stabilization of the 7p_1/2 AO with respect to the 6p_1/2 AO (Pershina, 2018; Pershina, 2016; Pershina et al., 2021). Thus, all the available predictions agree that both Nh and Mc should be more reactive than Cn and Fl and should easily adsorb at room temperature on Au, and also adsorb on less reactive quartz surfaces (Ilias and Pershina, 2022).

The gas chromatography method has proven being efficient for chemical studies of SHE compared to methods applied for chemical studies in the liquid phase (Türler and Pershina, 2013; Türler et al., 2015). However, the overall efficiency of this method is drastically dependent on the volatility and chemical reactivity of the species under study. In chemistry experiments with the lighter SHE rutherfordium (Rf, element 104) to bohrium (Bh, element 107), the chemical separation was achieved by the combination of the gas-jet transport with gas-solid chromatography (Türler and Pershina, 2013; Türler et al., 2015). A rather low overall efficiency of this multi-step technique was compensated by the relatively high production rates (Türler and Pershina, 2013). A significantly higher overall efficiency was achieved in the chemical study of hassium (Hs, element 108), which was separated in the form of the volatile compound HsO₄ (Düllmann et al., 2002). This technique is based on the combination of gas-solid isothermal chromatography and thermochromatography with a detection system, i.e., by registration of the nuclear decay of a species adsorbed inside the chromatography column itself. The method was developed further during the past two decades, all along to the chemistry studies of the rather inert Cn and Fl (Eichler et al., 2007; Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022). The necessity of the spatial separation of the single volatile superheavy atoms from other volatile species that contain radionuclides with similar nuclear decay properties as the nuclide under study and that are created in vastly larger amounts in the form of unwanted side products of the nuclear reactions was recognised (Düllmann et al., 2005). This includes especially Rn isotopes and their daughters. This separation is nowadays routinely achieved by employing physical preseparation in an electromagnetic recoil separator (Yakushev et al., 2014; Yakushev et al., 2022; Wittwer et al., 2010).

Several offline and online studies of the adsorption of Tl species on quartz and gold surfaces were performed, in which a stronger interaction of Tl with Au than with quartz was found (Serov et al., 2013; Steinegger et al., 2016; Lens et al., 2018). In thermochromatography studies under dry and oxygen-free conditions, as well as in the presence of hydrogen, the deposition of Tl on gold was observed at a temperature of about 900°C and was assigned to elemental Tl (Serov et al., 2013). However, independent of the chemical composition in the gas phase, the deposition of Tl on quartz at a temperature of about 300°C was ascribed to be due to TlOH formation caused by a high reactivity of Tl towards hydroxyl groups expected to be present on the quartz surface (Serov et al., 2013). Whereas initial experimental studies with homologues are usually performed in chromatography columns made of bulk materials like quartz (fused silica) or Au metal foils, SHE experiments are often conducted in channels made of silicon radiation detectors with thin hetero-layers covering the detector surfaces. Thin silicon oxide (SiO₂) or Au layers on silicon detectors are produced by oxidation or metal vapour deposition, respectively, and have poorly defined structures. These thin layers serve as reactive solid surface in the gas-solid chromatography process and are usually complex and differ from those used in most theoretical calculations on the adsorption of SHE.

For Cn and Fl, the focus of the chemical studies was on the interaction with Au surfaces (Eichler et al., 2007; Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022; Wittwer et al., 2010). In these experiments the adsorption enthalpy, − ∆ H ads Au , could be measured in a range of approximately 65–35 kJ/mol. Despite the limited number of observed Cn and Fl atoms, both elements were conclusively shown to exhibit high volatility and a rather low reactivity towards Au, characterizing them as volatile metals. In one specific experiment, a SiO₂-covered detector array was installed in front of the Au-covered ones. Neither Fl nor its decay product Cn adsorbed on this SiO₂ surface at room temperature (Yakushev et al., 2022). The low reactivity of Fl, which was confirmed experimentally in several experiments (Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022), is ascribed to the strong relativistic stabilization of the closed 7p_1/2 subshell (Schwerdtfeger et al., 2020; Pyper, 2020; Pershina, 2019). In contrast to rather inert Cn and Fl, which can be transported as gaseous species to a chromatography and detection setup (Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022), Nh and Mc might not. This makes experimental studies of these expectedly more reactive elements even more challenging (Yakushev et al., 2021). In fact, the transport of the presumably less volatile Nh to a detection setup was found to be inefficient (likely due to adsorption losses) through a tube made from a material as inert as polytetrafluoroethylene (PTFE) (Yakushev et al., 2021).

The longest-lived known Mc and Nh isotopes can be produced directly or as decay products in nuclear fusion reactions of ⁴⁸Ca with ²⁴³Am and ²⁴⁹Bk targets, leading to the production of Mc and tennessine (Ts, element 117) isotopes, respectively (Oganessian and Utyonkov, 2015). The experimentally easiest access to Mc and Nh is via the nuclear fusion reaction ²⁴³Am (⁴⁸Ca, 3n) ²⁸⁸Mc, which has a high cross section of approximately 10–20 pb (Oganessian et al., 2013; Rudolph et al., 2013; Gates et al., 2015; Oganessian et al., 2022; Oganessian et al., 2022b), similar to that for ²⁸⁸Fl produced in the reaction ²⁴⁴Pu (⁴⁸Ca, 4n) ²⁸⁸Fl (Düllmann et al., 2010; Såmark-Roth et al., 2023), which was successfully used in chemical studies (Eichler et al., 2010; Yakushev et al., 2014; Yakushev and Eichler, 2016; Yakushev et al., 2022). The decay properties of the members of the ²⁸⁸Mc decay chain are rather well established based on more than two hundred observed decay chains (Oganessian et al., 2013; Rudolph et al., 2013; Gates et al., 2015; Oganessian et al., 2022; Oganessian et al., 2022b). The second member of the ²⁸⁸Mc decay chain, ²⁸⁴Nh, has a half-life of T_1/2 = 0.90 − 0.06 + 0.07 s (Oganessian et al., 2022b), similar to those of the isotopes ^287-289Fl, which were in the focus of recent chemical studies (Yakushev et al., 2021; Eichler, 2013; Dmitriev et al., 2014; Aksenov et al., 2017). However, its directly produced shorter-lived precursor, ²⁸⁸Mc (T_1/2 = 193 − 13 + 15 ms) (Oganessian et al., 2022b), was out of reach due to the short lifetime that elapses before the Mc atoms reach the combined chromatography and detection system. The observation of several decay chains originating from ²⁸⁴Nh was claimed in early studies without physical pre-separation (Türler et al., 2015; Eichler, 2013; Dmitriev et al., 2014). The quality of the nuclear decay data in those works was limited by a rather high background in the α-particle spectra (Türler et al., 2015; Eichler, 2013; Dmitriev et al., 2014), which hampered the safe identification of decay chains (Türler et al., 2015; Yakushev et al., 2021). Two recent experiments employing preseparation were conducted behind the Dubna Gas-Filled Recoil Separator (DGFRS) (Akseneov et al., 2017), and behind the TransActinide Separator and Chemistry Apparatus (TASCA) (Yakushev et al., 2021). The low-background conditions led to a much higher sensitivity. However, no conclusive information on the chemical properties of Nh could be derived due to non-observation of ²⁸⁴Nh events in these experiments (Yakushev et al., 2021; Aksenov et al., 2017). The non-observation of Nh atoms in the experiments with preseparation was explained by losses due to the irreversible adsorption of Nh atoms on surfaces they encountered prior to reaching the detection setup, indicating a rather high chemical reactivity of the Nh atoms (Yakushev et al., 2021; Aksenov et al., 2017). Thus, no conclusive and confirmed results on the chemical properties of Nh were obtained in the previous studies. This called for further developments of the experimental setup. Gas chromatography studies of reactive species came within reach by employing the newly developed miniCOMPACT detector array (Yakushev et al., 2021), as demonstrated in adsorption studies of the lighter homologue of Mc, ²¹¹Bi ¹ and short-lived radioisotopes of the very reactive Fr (Götz et al., 2021). This improved detection system is promising for studies of Nh and is proven to be fast enough to observe a relevant fraction of all produced ²⁸⁸Mc.

The nuclear reaction ²⁴³Am (⁴⁸Ca, 3n)²⁸⁸Mc, utilized for the discovery of element 115, Mc (Oganessian and Utyonkov, 2015), and for the detailed studies of the decay properties of the long α-decay chain starting from the isotope ²⁸⁸Mc (Oganessian et al., 2013; Rudolph et al., 2013; Gates et al., 2015; Oganessian et al., 2022; Oganessian et al., 2022b), was also used for the present chemical studies of Mc and Nh. The gas-filled recoil separator TASCA (Semchenkov et al., 2008) was used to separate products of the fusion-evaporation reaction from the intense primary beam and from the unwanted products of multi-nucleon transfer reactions. The chemical reactivity of Mc and Nh was investigated applying the gas-solid chromatography method in the isothermal regime at room temperature (20°C) (Türler and Pershina, 2013; Türler et al., 2015) via adsorption studies on two surface types, SiO₂ and Au. The gas chromatography setup behind TASCA, as shown in Figure 1, was similar to that described in the first experiment on Nh chemistry at TASCA (Yakushev et al., 2021), except for the connecting tube between the Recoil Transfer Chamber (RTC) and the first detector array. The distance from the RTC gas volume to the first detector element was reduced to about 1 mm.

A pulsed (5 ms beam-on, 15 ms beam-off) ⁴⁸Ca¹⁰⁺ ion beam was provided by the UNIversal Linear ACcelerator (UNILAC) at GSI Darmstadt with a typical beam intensity of 5 10¹² s⁻¹. The target wheel contained four ²⁴³Am₂O₃ targets and rotated with a speed of 2000 revolutions per second, synchronized with beam pulses (Jäger et al., 2014). The ²⁴³Am targets with an average thickness of 0.83(1) mg/cm² were electro-deposited on 2.2(1) μm Ti foils (Runke et al., 2014). The beam energy in the middle of the ²⁴³Am targets was 243.0(2) MeV. In total, 1.2(1) 10^{19 48}Ca ions impinged on the targets during two experimental runs of 15 and 26 days duration, respectively. The evaporation residues were guided through the magnetic recoil separator TASCA with one dipole (D) and two quadrupole (Q) magnets in DQ₁Q₂ configuration, operated in the High-Transmission Mode (Semchenkov et al., 2008). TASCA was filled with He at a pressure of p _He = 0.8 mbar and was set to a magnetic rigidity of B·ρ = 2.21 Tm (Rudolph et al., 2013; Khuyagbaatar et al., 2012) to center ²⁸⁸Mc recoils in the TASCA focal plane. The nominal transmission efficiency in TASCA to focus the recoiling ²⁸⁸Mc ions into the RTC window area is about 40% (Yakushev et al., 2021). Products recoiling from the target and separated in TASCA according to their magnetic rigidities were thermalized in the RTC in a 1:1 gas mixture of He and Ar at 1 bar, similarly to the first Nh study at TASCA (Yakushev et al., 2021). Due to a large number of collisions with gas atoms, the isolated ions lose their kinetic energy, and their charge state is decreased accordingly to +1 or +2. The thermalized ions can be neutralized in gas by recombination or collisions with gas impurities or with any surface. The gas was circulated in a closed loop by a membrane pump at a flow rate of 2.9 L/min. The rare gases, He (99.9999%) and Ar (99.999%), were purified further by passing cartridges MC50-902FV and MC400-902FV (SAES™). These cartridges had internal particle filters, which efficiently remove aerosol particles with sizes down to 3 nm diameter and reduced the water and oxygen content to a level of about 1 ppm. The most critical impurity, water, was monitored by the dew point transmitter Pura (Michell Instruments). The isolated nuclear reaction products were flushed into the gas chromatography and detection setup. Two different detector channels were used, namely, a miniCOMPACT detector (Yakushev et al., 2021; Götz et al., 2021), followed by a COMPACT detector, used in the previous Fl chemistry studies at TASCA (Yakushev et al., 2014; Yakushev et al., 2022). The miniCOMPACT detector array was directly connected to the RTC via an exit slit with a cross section of (10.0 × 0.5) mm², fitting to the cross section of the miniCOMPACT channel, and allowing for an effective transport and detection of less-volatile species. As Nh and Mc should interact more strongly with gold than with quartz (Pershina et al., 2021), the silicon photodiodes of the miniCOMPACT detector array were covered with a 30–50 nm-thick SiO₂ layer. The photodiodes of the COMPACT detector array were covered with a 30 to 50 nm-thick Au layer. A 30-cm long capillary made of PTFE with an inner diameter of 2 mm connected the Au-covered COMPACT detector array to the miniCOMPACT array. Both detectors consisted of two panels, each divided into 32 detector elements. The opposite detector panels (top and bottom) were mounted with the active detector surfaces facing each other, and thus, forming a narrow gas channel with a distance between the opposite panels of 0.6 mm or 0.8 mm for the COMPACT array and the miniCOMPACT array, respectively. Each detector panel of the miniCOMPACT and COMPACT detectors consists of 32 Positive-Intrinsic-Negative (PIN) diodes mounted in a row (Wegrzecki et al., 2013). They registered α particles and fission fragments emitted by nuclear decays occurring inside the detector channel. Two different segmentation geometries of the miniCOMPACT detector arrays were used in the first and the second run. In the first run, a 15-cm long miniCOMPACT array had differently-sized detector elements mounted on a ceramic printed circuit board (PCB) as shown in Figure 2A. Each detector panel consisted of two silicon chips, which were 7 cm and 8 cm long. The first 7-cm long silicon chip was divided into three eight-element groups. They were 1, 2, and 4 cm long. The following 8-cm long chip had eight equal detector elements. In the second run, a 16-cm long miniCOMPACT detector array was used. The panels of the second detector consisted of two equal, 8-cm long silicon chips, each divided into sixteen detector elements, and were mounted on a multilayer FR4 PCB (flame retarding printed circuit board) as shown in Figure 2B.

Theoretical calculations accompany our experimental study. We have calculated electronic energies of gas-phase reactions that single metal atoms (M, where M = Bi and Mc) and singly-charged metal ions (M+) could undergo in the recoil chamber. The aim was to investigate a potential reactivity of Bi and Mc species towards water and oxygen, which are present as gas impurities at the trace level.

Our considered systems are SHE, which can only be treated appropriately when relativistic effects are explicitly taken into account in the calculations. A well-proven method for such work is the Density Functional Theory (DFT) approach (Hohenberg and Kohn, 1964). The theory underlying is the DFT is the self-consistent Kohn–Sham approach (Kohn and Sham, 1965). Kohn-Sham DFT is a first-principles computational method to predict accurately chemical properties and to analyse them in simple chemical terms. Specifically, the Kohn-Sham approach in general delivers correct electron densities and related properties, as well as the total energy, but implies a one-electron representation of the many-electron systems. Although the precise exchange-correlation (XC) functional is unknown, by using appropriate XC functionals, binding energies may typically be determined with a precision of a few kJ/mol (Te Velde et al., 2001). DFT keeps at all levels of approximation the appealing one-electron molecular orbital (MO) view on chemical reactions and properties. The computed orbitals are suitable for typical MO-theoretical analyses and interpretations. The Kohn-Sham method also effectively incorporates all correlation effects. The calculations of electronic structures of the molecules of interest were performed using the Density Functional Theory approach implemented in the AMS ADF software (SCM, 2024). The AMS electronic structure approach uses the zeroth order regular approximation (ZORA) two-component Hamiltonian incorporating both scalar relativistic (SR) and SO relativistic effects. Although the regular approximation is not gauge invariant at any order, the scaled ZORA energy is exactly gauge invariant for hydrogen ions. The scaled ZORA approach, as it was used in our ADF calculations, is nearly gauge-invariant for many-electron systems, as explained in detail in Ref. (Van Lenthe et al., 1994). Briefly, the scaled ZORA equation differs from the original ZORA equation by the denominator to introduce a scaling of the ZORA energy. The calculated scaled-ZORA total energies prove to be very close to the fully relativistic energies for atoms. For all-electron calculations, the original ZORA total energy cannot be used as it suffers from gauge invariance problems that are particularly serious for deep-core states. The scaled ZORA total energy solved this problem, delivering results within “chemical” accuracy. The same holds for two-component calculations with the SO ZORA Hamiltonian. The spin-polarized non-collinear approach, also called as Kramers unrestricted approach, was applied. The basis sets are a combination of numerical AOs and Slater-type orbitals (STOs). The used basis sets are of the Triple Zeta plus Double Polarization (TZ2P) size. These basis sets are well saturated and describe the polarization of the electron density of an atom in molecules. Since correlation effects are also very important for SHE systems and can contribute up to the 50% of the binding energy, those are taken into account by using XC functionals. We utilized the BP86 (Becke, 1988) exchange-correlation functional. The geometry optimizations of studied molecules were carried out with the SO ZORA Hamiltonian, ensuring that both SR and SO relativistic effects were reflected in the optimized molecular structures.

A search for position- and time-correlated nuclear decay chains was performed, to identify multiple genetically-linked members of the well-known long ²⁸⁸Mc decay chain, decaying by emission of α-particles within the energy interval (8.0–10.5) MeV and within a 200-s time interval. This search resulted in the finding of eighteen decay chains, seven in the first run and eleven in the second run. The decay chains were assigned to ²⁸⁸Mc or to ²⁸⁴Nh in accordance with their known decay properties (Oganessian et al., 2022; Oganessian et al., 2022b; Forsberg et al., 2016; Rudolph, 2022). As usual for such an experiment, where an atom is adsorbed on a detector surface inside a gas chromatography channel, α particles or fission fragments have to penetrate several inactive layers. They deposit a fraction of their kinetic energy in a SiO₂ or Au layer on top of the detector surface, and then in an inactive silicon layer on top of the diode. In addition, if the particles cross the detector channel, they experience further energy loss in the gas. These energy losses depend strongly on the angle to the surface, at which the particle is emitted. Thus, the energy distributions of registered α particles and spontaneous fission (SF) fragments have usually tailing towards low energies in contrast to the orresponding energy distributions registered in physics experiments, where the ions are implanted into a silicon detector. The energy calibration of the detectors was performed using a ²²⁷Ac source emanating ²¹⁹Rn, and was linearly extrapolated up to the fission fragment energies. Independently, the search for high-energy events with an energy of E_SF > 20 MeV was performed on a time scale of hours to days, to identify candidates for long-lived decay chain members terminating the decay chains by SF. In total, twenty-one candidate SF events were identified in the two runs. Sixteen SF events, twelve with two fragments and four with one fragment, occurred at the same detector positions, where chains with multiple α-decaying members were also detected, several hours before the SF events. These position-correlated SF events were assigned to the SF decay of ²⁶⁸Db or ²⁶⁴Lr (Oganessian et al., 2022). For two out of the eighteen registered decay chains, no terminating SF events were found, presumably due to a several-hour long problem with the data acquisition system in the period following the detection of these two chains. The decay parameters, energies and time intervals of the members of all observed decay chains are listed in Table 1. Figure 3 shows four representative examples (b–e) of observed chains, along with the reference decay chain (a).

Based on the decay properties along the long ²⁸⁸Mc decay chain (Figure 3A), four out of eighteen chains were attributed to originate from ²⁸⁸Mc (cf. Table 1). The missing observation of ²⁸⁸Mc in the other fourteen chains is attributed to the decay losses of the short-lived nucleus ²⁸⁸Mc ( T 1 / 2 = 193 − 13 + 15 ms) during the transport of the ²⁸⁸Mc atoms into the detector channel. The observed ratio of the number of detected decay chains starting with ²⁸⁸Mc decay to that of chains starting with ²⁸⁴Nh (T_1/2 = 0.90 − 0.06 + 0.07 s) allows a rough estimate of the mean flush-out time from the RTC window into miniCOMPACT of about 0.4 s. The transport time corresponds to approximately two half-lives for ²⁸⁸Mc, but it was short enough for the efficient transport of ²⁸⁴Nh. Within the observed decay chains, eleven α particles from the decay chain members ²⁸⁴Nh, ²⁸⁰Rg, ²⁷⁶Mt, ²⁷⁰Bh, were missing, in agreement with the mean detection efficiency for a single α particle emitted inside the miniCOMPACT detector channel of about 80%. The search for the α decay of the long-lived ²⁶⁸Db was performed within the energy interval (7.7–8.0) MeV and within the time interval between the last member of the α-decay chain occurring during the time interval of 200 s and the concluding SF event. Several candidates for α decay of ²⁶⁸Db were found (Table 1). However, this search was hindered in some cases by a non-zero background in this energy range, especially from non-volatile byproducts within the first centimeters of the miniCOMPACT detector. The measured time intervals between neighbouring members of all registered decay chains show a very good agreement with the known half-lives, deduced from the observation of about two hundred chains in physics experiments (Oganessian et al., 2013; Rudolph et al., 2013; Gates et al., 2015; Oganessian et al., 2022b). The experimental distributions of correlation times for the decay chain members ²⁸⁴Nh, ²⁸⁰Rg, ²⁷⁶Mt, ²⁷²Bh and ²⁶⁸Db are shown in Figure 4, in comparison with lifetime probability curves calculated for the recently published most precise half-lives (Oganessian et al., 2022; Oganessian et al., 2022b).

Five SF events, which were observed without position and time correlation to any α-decaying precursors within the energy range (9.0–10.5) MeV and a time interval of 1000 s. They could not be assigned to any registered long decay chain from ²⁸⁸Mc or ²⁸⁴Nh. Two such SF events were registered in the first centimeter of the miniCOMPACT array, and three at the 4th, 6th and 10th centimeter. In four cases, both fission fragments were detected. These SF-only events were considered as real ones and could be attributed to SF branches in early chain members (Oganessian et al., 2022; Forsberg et al., 2016; Rudolph, 2022). A detailed analysis on decay scenario aspects along the decay chain from ²⁸⁸Mc will be published elsewhere ² . Even though the spatial distribution of these SF events in the detector channel resembles the distribution of decay chains from ²⁸⁴Nh, the ambiguity in the assignment of these SF-only events did not allow including them for the evaluation of the chemical behaviour.

A comparison of the number of events registered in the present experiment to the number of decay chains from the Mc decay spectroscopy experiment at TASCA (Rudolph et al., 2013) normalized to the number of produced ²⁸⁸Mc nuclei reveals that roughly 40% of the reaction products guided to the TASCA focal plane were extracted into the gas chromatography channel. The losses were primarily caused by adsorption of non-volatile species on the RTC surface before reaching the detection setup. In addition, decay losses also significantly contributed to the overall losses of the short-lived ²⁸⁸Mc.

All observed decay chains originating from ²⁸⁸Mc or ²⁸⁴Nh were found in the miniCOMPACT detector, except for one decay chain from ²⁸⁴Nh, which was registered in the first centimeter of the Au-covered COMPACT. Thus, while all four Mc atoms adsorbed on the SiO₂ surface, one of the 14 Nh atoms passed the miniCOMPACT detector and a 30-cm long connecting tube made of PTFE between miniCOMPACT and COMPACT, but adsorbed on the very first Au-covered detector. This observation may suggest a stronger interaction with the Au surface compared to interaction with SiO₂ and PTFE surfaces, in line with the predictions (Pershina et al., 2021). This encourages future studies with higher statistics. The positions of the observed decay chains starting with ²⁸⁸Mc and starting with ²⁸⁴Nh in the detector channel are presented in Figure 5. All four events originating from ²⁸⁸Mc and about 85% of the ²⁸⁴Nh events were registered in the first half of the SiO₂-covered miniCOMPACT detector.