The periodic table provides a classification of the chemical properties of the elements. But for the heaviest elements, the transactinides, this role of the periodic table reaches its limits because increasingly strong relativistic effects on the valence electron shells can induce deviations from known trends in chemical properties1,2,3,4. In the case of the first two transactinides, elements 104 and 105, relativistic effects do indeed influence their chemical properties5, whereas elements 106 and 107 both behave as expected from their position within the periodic table6,7. Here we report the chemical separation and characterization of only seven detected atoms of element 108 (hassium, Hs), which were generated as isotopes 269Hs (refs 8, 9) and 270Hs (ref. 10) in the fusion reaction between 26Mg and 248Cm. The hassium atoms are immediately oxidized to a highly volatile oxide, presumably HsO4, for which we determine an enthalpy of adsorption on our detector surface that is comparable to the adsorption enthalpy determined under identical conditions for the osmium oxide OsO4. These results provide evidence that the chemical properties of hassium and its lighter homologue osmium are similar, thus confirming that hassium exhibits properties as expected from its position in group 8 of the periodic table.


The discovery of Hs was reported in 1984 (ref. 11) with the identification of the nuclide 265Hs with a half-life of T1/2 = 1.5 ms (refs 11, 12). In 1996, the much longer-lived isotope 269Hs, with a half-life of about 10 s, was observed in the α-decay chain of 277112 (ref. 8). Recently, evidence for the existence of the neighbouring nuclide 270Hs was found and a half-life of about 4 s was deduced from its measured α-decay energy10. The latter two Hs isotopes might be sufficiently long-lived to allow their chemical characterization. Suitable reactions for their direct production are 248Cm(26Mg, 5,4n)269,270Hs, for which formation cross-sections of a few picobarn have been estimated13.

The periodic table suggests that Hs is a member of group 8 and thus chemically similar to its lighter homologues ruthenium (Ru) and osmium (Os), which are known to form highly volatile tetroxides. Therefore, Hs is also expected to form a very volatile tetroxide (HsO4) suitable for gas-phase isolation1,14,15,16, even though two earlier attempts to chemically identify Hs in the tetroxide form proved unsuccessful17,18.

Fully relativistic density functional calculations19 for the tetroxides of the group 8 elements indicated that the electronic structure of HsO4 is very similar to that of OsO4, with covalent bonding being somewhat more pronounced in the former. The stability of the gaseous tetroxides was found19 to increase in the order RuO4 < OsO4 < HsO4, in agreement with extrapolations within group 8 of the periodic table20. The density functional calculations, in conjunction with a surface interaction model, suggest adsorption enthalpies of HsO4 and OsO4 on a quartz surface of (-35.9 ± 1.5) kJ mol-1 and (-38.0 ± 1.5) kJ mol-1, respectively. Extrapolations of the volatility of group 8 tetroxides suggest almost identical adsorption enthalpies of (- 46 ± 15) kJ mol-1 and (-45 ± 15) kJ mol-1 for HsO4 and OsO4, respectively, whereas a physisorption model yields identical values of (-47 ± 11) kJ mol-1 for both tetroxides. Overall, the theoretical values suggest fairly similar adsorption behaviour for HsO4 and OsO4 on quartz. Although the present experiment uses detectors covered with a silicon nitride layer, we expect a direct analogy between theory and experiment because we observed comparable adsorption interactions of OsO4 with quartz and silicon nitride.

The expected high volatility of group 8 tetroxides allows excellent separation from heavy actinides and lighter transactinides as well as from Pb, Bi and Po, which is important because several isotopes of these elements are formed with high yield as byproducts of the nuclear fusion reaction. These nuclides often decay via emission of α-particles and, therefore, severely interfere with the unambiguous detection of the nuclear decay of the investigated Hs nuclide.

Online investigations of Os oxides have already been conducted earlier15,21,22. Using the in situ volatilization and online detection apparatus (IVO) carrier-free 173OsO4 (T1/2 = 22.4 s) could be separated and detected with an overall efficiency of (40 ± 10)%. Decontamination from Po (≥2 × 104) was excellent21.

The experimental set-up, schematically shown in Fig. 1, involves mounting the 248Cm target23 on a rotating wheel and bombarding it with up to 8 × 1012 26Mg5+ particles per second, delivered by the UNILAC accelerator at the Gesellschaft für Schwerionenforschung mbH. The particle beam first passed through a rotating three-segment 3.68 mg cm-2 Be vacuum window which allowed for a pressure up to 1.3 atm in the recoil chamber. The 192.7-MeV beam energy delivered by the UNILAC resulted in 26Mg projectile energies of 143.7–146.8 MeV inside the 248Cm target. Nuclear reaction products recoiling from the target were thermalized in the gas volume (34 ml) of the IVO device21 flushed with dry (measured water vapour concentration <1 p.p.m.) 1.2 l min-1 helium (He) and 100 ml min-1 oxygen (O2). The reaction products were transported with the carrier gas through a 30-cm-long quartz column (internal diameter, i.d., 4 mm) containing a quartz wool plug at a distance of 6.5 cm from the recoil chamber. This plug was heated to 600 °C and served as a filter for aerosol particles and provided a surface to complete the oxidation reaction of Os and Hs to their tetroxides, which were further transported through a 10-m-long perfluoroalkoxy (PFA) Teflon capillary (i.d., 2 mm) to the detection system. The capillary was installed inside a polyamide (PA) tube flushed with dry N2.

Figure 1: Schematic drawing of the IVO-COLD set-up used to produce and isolate Hs isotopes in form of the volatile HsO4.
Figure 1

The 26Mg-beam (1) passes through the rotating vacuum window and 248Cm-target (2) assembly. The target consisted of three banana-shaped segments (1.9 cm2 area each) covered with 239 µg cm-2, 730 µg cm-2, and 692 µg cm-2 248Cm, respectively. The 248Cm (isotopic composition 246Cm: 4.2%; 248Cm: 95.8%) was deposited on 2.82 mg cm-2 beryllium (Be) backings by molecular plating. The target-window assembly rotated in the adjacent gas volume with 2,000 rev min-1 and was synchronized with the beam macrostructure of the accelerator in order to distribute each 6-ms beam pulse evenly over one target segment. In the fusion reaction 269,270Hs nuclei are formed that recoil out of the target into a gas volume (3) and are flushed with a He/O2 mixture (4) out of the chamber. The gas was passed through a cartridge containing P2O5 as a drying agent before injecting it into IVO. In this way, the water vapour concentration was reduced to a measured value of <1 p.p.m. throughout the experiment. The gas was injected into a quartz column (5) containing a quartz wool plug (6) heated to 600 °C by an oven (7). There, Hs is converted to HsO4 which is volatile at room temperature and transported with the gas flow through a 10-m-long perfluoroalkoxy (PFA) Teflon capillary (8) to the COLD detector array registering the nuclear decay (α and spontaneous fission) of the Hs nuclides. The array consists of 24 detectors arranged in 12 pairs (9). A temperature gradient was established along the detector array by means of a thermostat (10) at the entrance and a liquid nitrogen cryostat (11) at the exit. The temperature was monitored by five thermocouples installed along the copper bar. Depending on the volatility of HsO4, the molecules adsorb at a characteristic temperature.

Using gas phase adsorption thermochromatography24, we measured the temperature at which HsO4 deposits, with a chromatographic column along which a stationary negative temperature gradient is maintained. The chromatographic column, the cryo online detector (COLD), also served as detection system for the identification of decaying atoms of 269,270Hs. COLD consists of 12 pairs of silicon PIN-photodiodes of 1 × 3 cm2 active area mounted at a distance of 1.5 mm via two spacers made from silicon. PIN diodes are well suited for detection of α-particles or fission fragments. The diode pairs were installed inside a copper bar. A temperature gradient from -20 °C to -170 °C was established along the detector array. The efficiency for detecting a single α-particle emitted by a species adsorbed within the detector array was 77%. To avoid formation of ice layers on the detector surfaces, the whole device was placed inside a vacuum tight housing flushed with dry N2. COLD is an improved version of a previous set-up called the cryo-thermochromatography separator (CTS)22.

The detectors of the COLD array were online calibrated with α-decaying 219Rn and its daughters 215Po and 211Bi using a 227Ac source. The determined energy resolution was 50–70 keV (full-width at half-maximum, FWHM) for detectors 1 through 8 and 80–110 keV (FWHM) for detectors 9 through 12.

The proper functioning of the IVO-COLD system was checked at the beginning and after the end of the Hs experiment by mounting a 800 µg cm-2 152Gd (32% enriched) target to produce short-lived Os isotopes in the reaction 152Gd(26Mg, 6n)172Os. The beam intensity was 1 × 1012 s-1 and the beam energy in the middle of the target 153 MeV. All other experimental parameters, including gas composition and flow rate, were identical to those of the Hs experiment. In the COLD detector array the α-decay of 172Os was registered. 172Os has a half-life of 19.2 s and a 1.0% α-decay branch with Eα = 5.10 MeV.

The experiment to produce Hs isotopes lasted 64.2 h during which a total of 1.0 × 1018 26Mg particles passed through the target. Only α-lines originating from 211At, 219,220Rn and their decay products were identified. Whereas 211At and its decay product 211Po were deposited mainly in the first two detectors, 219,220Rn and their decay products accumulated in the last three detectors, where the temperature was sufficiently low to partly adsorb Rn. One side of detector pair 1 did not operate owing to a technical failure and, therefore, this detector unit was excluded from the data analysis. During the experiment, seven correlated decay chains were detected (see Fig. 2). All correlated α-decay chains were observed in detectors 2 through 4 and assigned to the decay of 269Hs or 270Hs. The characteristics of the first three decay chains agree well with data8,9 on 269Hs and its daughter nuclides, while two other decay chains can be attributed10 to the decay of 270Hs. The last two decay chains were incomplete and a definite assignment to 269Hs or 270Hs could not be made. No additional three-member decay chains within ≤300 s were registered in detectors 2 to 10. The background count-rate of α-particles with energies between 8.0 and 9.5 MeV was about 0.6 h-1 per detector, leading to very low probabilities of ≤7 × 10-5 and ≤2 × 10-3 for any of the first five chains and any of the last two chains, respectively, being of random origin. In addition, four uncorrelated fission fragments with energies >50 MeV were registered in detectors 2 through 4. All other detectors registered no fission fragments, except for one fission fragment being observed in the operating side of detector 1.

Figure 2: The seven nuclear decay chains originating from Hs isotopes that were detected in the course of the experiment permit an unambiguous identification of hassium after chemical separation.
Figure 2

Indicated are the energies of α-particles and fission fragments in megaelectronvolts (MeV) and the lifetimes in seconds (s). The detector in which the decay was registered is indicated in parentheses where T stands for top detector and B for bottom detector. For each chain the date and time of its registration are given. The lifetime of the mother isotope could not be determined with the applied thermochromatography technique because the deposition time is not measured.

As depicted in Fig. 3, the α-decay of one Hs atom was registered in detector 2, the decay of four atoms in detector 3 and the decay of two atoms in detector 4. The maximum of the Hs distribution was found for a temperature of (-44 ± 6) °C. The deposition distribution of OsO4 measured before and after the experiment revealed a maximum in detector 6 at (-82 ± 7) °C.

Figure 3: Merged thermochromatogram of HsO4 and OsO4.
Figure 3

Indicated are the relative yields of HsO4 and OsO4 for each of the 12 detector pairs. Measured values are represented by bars: HsO4, dark grey; OsO4, white. For Hs, the following distribution was measured: 269Hs: three events (all in detector 3); 270Hs: two events (one in detector 2 and one in detector 4); Hs (isotope unknown): two events (one in detector 3 and one in detector 4). For Os, the distribution of 1 × 105 events of 172OsO4 is given. The dashed line indicates the temperature profile (right-hand scale). The maxima of the deposition distributions were evaluated as (-44 ± 6) °C for HsO4 and (-82 ± 7) °C for OsO4 where the uncertainties indicate the temperature range covered by the detector which registered the maximum of the deposition distribution. Solid lines represent results of a Monte Carlo simulation of the migration process of the species along the column with standard adsorption enthalpies of -46.0 kJ mol-1 for 269HsO4 and -39.0 kJ mol-1 for 172OsO4.

The adsorption enthalpy (ΔHads) of the compound on the stationary phase is extracted from the measured deposition distribution by using Monte Carlo simulations of the trajectories of single molecules as they move along the column under real experimental conditions25. The only free parameter in the simulations is ΔHads, but the half-life of the nuclide is a crucial parameter. For this reason, and because the half-life of 270Hs has not yet been measured, only decays assigned to 269Hs were used to evaluate the adsorption enthalpy of the compound on the silicon nitride surface. The results that best reproduce the experimental data are shown in Fig. 3 (solid lines) and suggest a value of ΔHads = (-46 ± 2) kJ mol-1 (68% confidence interval, c.i.), which was inferred using a T1/2 value of 11+15-4 s for 269Hs (refs 8, 9). The given uncertainty is the total uncertainty; that is, it reflects the width of the measured deposition peak and the uncertainty in T1/2. The adsorption enthalpy of OsO4 on silicon nitride deduced from this experiment was (-39 ± 1) kJ mol-1, which is in good agreement with the adsorption enthalpy obtained in earlier investigations using quartz surfaces15,21,22.

The experimentally derived adsorption enthalpy of HsO4 is thus lower than that of OsO4, while the predictions suggest either similar values20 or a slightly higher value19 for HsO4. However, the theoretical values and experimental values have associated uncertainties. Moreover, the low value of the ΔHads determined for the hassium oxide species clearly suggests that it is HsO4, given that, by analogy with the known properties of the Os oxides, other Hs oxides are all expected to be less volatile and unable to reach the detection system. The observed formation of a very volatile Hs molecule, presumably HsO4, in a mixture of oxygen and helium thus provides strong qualitative evidence that Hs is an ordinary member of group 8 of the periodic table that behaves similarly to its lighter homologue Os.


  1. 1.

    Superheavy elements. Struct. Bonding 21, 90–144 (1975)

  2. 2.

    & Relativity and the periodic system of elements. Acc. Chem. Res. 12, 276–281 (1979)

  3. 3.

    Electronic structure and properties of the transactinides and their compounds. Chem. Rev. 96, 1977–2010 (1996)

  4. 4.

    & Relativistic effects of the superheavy elements. Encyclopedia of Computational Chemistry Vol. 4 2480–2499 (Wiley, New York, 1998)

  5. 5.

    in Heavy Elements and Related New Phenomena Ch. 4 (eds Greiner, W. & Gupta, R. K.) 129–193 (World Scientific, Singapore, 1999)

  6. 6.

    et al. Chemical properties of element 106 (seaborgium). Nature 388, 55–57 (1997)

  7. 7.

    et al. Chemical characterization of bohrium (element 107). Nature 407, 63–65 (2000)

  8. 8.

    et al. The new element 112. Z. Phys. A 354, 229–230 (1996)

  9. 9.

    & The discovery of the heaviest elements. Rev. Mod. Phys. 72, 733–767 (2000)

  10. 10.

    et al. Decay properties of 269Hs and evidence for the new nuclide 270Hs. Eur. Phys. J. A (submitted)

  11. 11.

    et al. The identification of element 108. Z. Phys. A 317, 235–236 (1984)

  12. 12.

    et al. Production and decay of 269110. Z. Phys. A 350, 277–280 (1995)

  13. 13.

    & Prospects for the discovery of new elements. J. Radioanal. Nucl. Chem. 203, 283–300 (1996)

  14. 14.

    & Chemische Probleme bei der Darstellung überschwerer Elemente durch Kernreaktionen. Radiochim. Acta 15, 153–163 (1971)

  15. 15.

    & Continuous-flow thermochromatographic separation of unsupported radioisotopes of platinum elements in a stream of air from nuclear reaction products in an accelerator heavy-ion beam. Sov. Radiochem. 26, 731–739 (1985); translated from Radiokhimiya 26, 770–778 (1984)

  16. 16.

    , , , & Thermochromatography of platinum elements in oxygen: Radiochemical studies of the behaviour of rhodium, palladium, osmium and platinum. Radiochim. Acta 62, 61–63 (1993)

  17. 17.

    Zhuikov, B. L., Kruz, H. & Zvara, I. Possibilities of chemical identification of short-lived isotopes of element 108. Report P7-86-322, page 26 (Joint Institute for Nuclear Research, JINR, Dubna, 1986) (in Russian).

  18. 18.

    Dougan, R. J., Moody, K. J., Hulet, E. K. & Bethune, G. R. OSCAR: An apparatus for on-line gas-phase separations. FY87 Annual Report UCAR 10062/87, 4–17 (Lawrence Livermore National Laboratory, LLNL, Nuclear Chemistry Division, Livermore, 1987).

  19. 19.

    , , & The electronic structure and properties of group 8 oxides MO4, where M = Ru, Os, and element 108, Hs. J. Chem. Phys. 115, 792–799 (2001)

  20. 20.

    , , , & On the stability and volatility of group 8 tetroxides MO4 (M = ruthenium, osmium, and hassium (Z = 108)). J. Phys. Chem. B 106, 6679–6684 (2002)

  21. 21.

    et al. IVO, a device for in situ volatilization and on-line detection of products from heavy ion reactions. Nucl. Instrum. Meth. A 479, 631–639 (2002)

  22. 22.

    et al. The cryo-thermochromatographic separator (CTS): A new rapid separation and α-detection system for on-line chemical studies of highly volatile osmium and hassium (Z = 108) tetroxides. Nucl. Instrum. Meth. A 484, 587–594 (2002)

  23. 23.

    et al. Separation of 248Cm from a 252Cf neutron source for production of Cm targets. Radiochim. Acta 89, 543–549 (2001)

  24. 24.

    Thermochromatographic method of separation of chemical elements in nuclear and radiochemistry. Isotopenpraxis 26, 251–258 (1990)

  25. 25.

    Simulation of thermochromatographic processes by the Monte Carlo method. Radiochim. Acta 38, 95–101 (1985)

Download references


We thank the staff of the Laboratory for Micro- and Nanotechnology at PSI for manufacturing the PIN-diode sandwiches for the COLD array and the staff of the GSI UNILAC for providing stable, highly intense beams of 26Mg as well as the target laboratory for Be foils for the vacuum windows. Support from the European Commission Institute for Transuranium Elements, Karlsruhe, for long-term storage of 252Cf and the chemical separation of 248Cm is appreciated. These studies were supported in part by the Swiss National Science Foundation and the Chemical Sciences Division of the Office of Basic Energy Sciences, US Department of Energy.

Author information


  1. *Departement für Chemie und Biochemie, Universität Bern, CH-3012 Bern, Switzerland

    • Ch. E. Düllmann
    • , H. W. Gäggeler
    •  & S. Soverna
  2. †Labor für Radio- und Umweltchemie, Paul Scherrer Institut, CH-5232 Villigen, Switzerland

    • Ch. E. Düllmann
    • , R. Dressler
    • , B. Eichler
    • , R. Eichler
    • , H. W. Gäggeler
    • , F. Glaus
    • , D. T. Jost
    • , D. Piguet
    •  & S. Soverna
  3. ‡Gesellschaft für Schwerionenforschung mbH, D-64291 Darmstadt, Germany

    • W. Brüchle
    • , E. Jäger
    • , V. Pershina
    • , M. Schädel
    • , B. Schausten
    • , E. Schimpf
    • , H.-J. Schött
    •  & G. Wirth
  4. §Institut für Kernchemie, Universität Mainz, D-55128 Mainz, Germany

    • K. Eberhardt
    • , P. Thörle
    •  & N. Trautmann
  5. Nuclear Science Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, USA

    • T. N. Ginter
    • , K. E. Gregorich
    • , D. C. Hoffman
    • , U. W. Kirbach
    • , D. M. Lee
    • , H. Nitsche
    • , J. B. Patin
    • , R. Sudowe
    •  & P. M. Zielinski
  6. ¶Department of Chemistry, University of California, Berkeley, California 94720-1460, USA

    • D. C. Hoffman
    • , H. Nitsche
    •  & J. B. Patin
  7. #Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou 730000, P.R. China

    • Z. Qin
  8. Flerov Laboratory of Nuclear Reactions, Joint Institute for Nuclear Research, 141980 Dubna, Russia

    • S. N. Timokhin
    •  & A. B. Yakushev
  9. **Institut für Radiochemie, Technische Universität München, D-85748 Garching, Germany

    • A. Türler
  10. ††Research Center Rossendorf e.V., D-01314 Dresden, Germany

    • A. Vahle


  1. Search for Ch. E. Düllmann in:

  2. Search for W. Brüchle in:

  3. Search for R. Dressler in:

  4. Search for K. Eberhardt in:

  5. Search for B. Eichler in:

  6. Search for R. Eichler in:

  7. Search for H. W. Gäggeler in:

  8. Search for T. N. Ginter in:

  9. Search for F. Glaus in:

  10. Search for K. E. Gregorich in:

  11. Search for D. C. Hoffman in:

  12. Search for E. Jäger in:

  13. Search for D. T. Jost in:

  14. Search for U. W. Kirbach in:

  15. Search for D. M. Lee in:

  16. Search for H. Nitsche in:

  17. Search for J. B. Patin in:

  18. Search for V. Pershina in:

  19. Search for D. Piguet in:

  20. Search for Z. Qin in:

  21. Search for M. Schädel in:

  22. Search for B. Schausten in:

  23. Search for E. Schimpf in:

  24. Search for H.-J. Schött in:

  25. Search for S. Soverna in:

  26. Search for R. Sudowe in:

  27. Search for P. Thörle in:

  28. Search for S. N. Timokhin in:

  29. Search for N. Trautmann in:

  30. Search for A. Türler in:

  31. Search for A. Vahle in:

  32. Search for G. Wirth in:

  33. Search for A. B. Yakushev in:

  34. Search for P. M. Zielinski in:

Competing interests

The authors declare that they have no competing financial interests.

Corresponding author

Correspondence to H. W. Gäggeler.

About this article

Publication history





Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.