Lifetime Measurement of the Metastable 2 3 P 0 State in Helium-like 238 U S. Toleikis 1 , B. Manil 2* , G. Bednarz 3** , E. Berdermann 1 , H.F. Beyer 1 , F. Bosch 1 , A. Bräuning-Demian 1 , A. Gumberidze 1 , S. Hagmann 1 , P. Indelicato 2* , C. Kozhuharov 1 , D. Liesen 1 , X. Ma 6 , R. Marrus 4 , P.H. Mokler 1 , Z. Stachura 5 , Th. Stöhlker 1 , A. Warczak 3 1 GSI, Darmstadt, Germany, 2 Université P. et M. Curie, Lab. Kastler Brossel, Paris, France, 3 IFUJ, Cracow, Poland, 4 University of California, Berkeley, U.S.A., 5 INP, Cracow, Poland, 6 IMP, Lanzhou, China The measurement we report here is the second of a series of lifetime measurements on helium-like heavy ions for investiga- tions of fundamental atomic properties. In the first experiment on helium-like 197 Au we were able to determine the lifetime of the hyperfine quenched 2 3 P 0 state with an accuracy of ~3% [1]. Due to the hyperfine interaction, the 197 Au nucleus has spin I =3/2, the metastable 2 3 P 0 state couples to the prompt 2 3 P 1 state hereby reducing its lifetime. As the coupling strength depends on the 2 3 P 0 -2 3 P 1 fine-structure splitting, one can de- termine this fine-structure splitting by measuring the lifetime of the 2 3 P 0 state. The situation in helium-like 238 U is fundamentally different because the 238 U nucleus has no nuclear spin. The consequence is that no hyperfine interaction occurs. The lifetime of the 2 3 P 0 state is therefore determined by the decay rate of the 2 3 P 0 state to the 2 3 S 1 state which itself decays to the ground state by a fast M1 transition (see Figure 1). As the decay rate, respectively the 1 S 2 S 2 S 2 P 2 P 2 P 2 P 0 1 1 1 3 1 2 3 3 0 1 3 0 1 96.285 keV 2E1 ~5(12) 100.614 keV E1 5.0(16) 4.585 keV E1 3.0(13) 4.511 keV E1 8.6(13) 0.110 keV 1.2(9) M1 0.253 keV E1 1.2(10) 100.540 keV M2 2.1(14) 96.030 keV M1 1.2(14) E1 96.173 keV E1M1 96.282 keV 5.6(9) 3.0(16) Figure 1: Level scheme of helium-like uranium with transition energies, decay modes and transition probabilities indicated. Numbers are taken from [2] and [5]. The transition probabil- ities are in 1/s with numbers in brackets indicating powers of 10. lifetime, mainly depends on the energy difference between the 2 3 P 0 and the 2 3 S 1 state which is in fact the n=2 Lamb shift, one can in principle deduce the Lamb shift from a lifetime measure- ment of the 2 3 P 0 state. A previous experiment at the Lawrence Berkeley Laboratory’s Bevalac resulted in a value of 54.4±3.4 ps for the lifetime of the 2 3 P 0 state [3]. The basic method for the measurement of the lifetime of metastable atomic states in heavy ions is beam foil spec- troscopy. Figure 2 shows the apparatus used in our experiment. Ge(i) MOVE dipole bare He-like H-like diamond detector (32 stripes) Cave A E=289.8 MeV/u β =0.6467 target foil for electron capture Ni, 1.5 mg/cm² x coincidence ion beam Ge(i) FIX collimators (Ta) H-like U uranium ions/s from SIS Figure 2: Experimental setup of the lifetime measurement in Cave A performed in May 2001. The different charge states of the uranium ions produced in the target foil are separated by a magnet spectrometer and are detected by a position sensitive diamond detector. A hydrogen-like uranium beam with an energy of 289.8 MeV/u (β =0.6467) passes through a target foil (1.5 mg/cm 2 Ni) and hereby produces excited helium-like ions by single elec- tron capture. The radiation of the subsequent decay of the ex- cited states is detected downstream of the foil by two Ge(i) de- tectors, located on opposite sides of the beam. The position of one detector is fixed while the other detector is moveable. By varying the distance between the target foil and the moveable detector and by measuring the ratio of counts of the 2 3 S 1 -1 1 S 0 - transition in the moveable detector relative to the fixed detector, a decay curve can be traced out. Measuring the ratio allows nor- malisation to the ion population in the excited state of interest and has the additional advantage that most systematic errors are eliminated. The fact that one detects the 2 3 S 1 -1 1 S 0 -transition and not directly the 2 3 P 0 -2 3 S 1 -transition has several reasons. First of all, the accurate detection of the direct 253 eV transition (see Figure 1) is not possible, because no detector with a suffi- cient energy resolution and efficiency is available in this energy region. Secondly, all other feeding mechanisms of the 2 3 S 1 state have died out long before the metastable 2 3 P 0 state starts to decay. A possible contamination of the line of interest (2 3 S 1 - 1 1 S 0 -transition) due to the exotic two-photon (E1M1) branch from the 2 3 P 0 state directly to the ground state can be neglected as these photons are almost uniformly distributed over the en- tire range from 0 to 96.3 keV. Therefore by detecting the 2 3 S 1 - 1 1 S 0 -transition and tracing the decay curve the lifetime of the 2 3 P 0 state can be determined. A raw sample x-ray spectrum of the moveable detector is shown in Figure 3. The problem of background radiation is ob-