VOLUME 69, NUMBER 7 PH YSICAL R EVI EW LETTERS 17 AUGUST 1992 Phase Effect in the Energy Loss of H Projectiles in Zn Targets: Experimental Evidence and Theoretical Explanation P. Bauer, ' ' F. Kastner, A. Arnau, ' A. Salin, P. D. Fainstein, V. H. Ponce, and P. M. Echenique ' ' Departamento de Fisica de Materiales, Universidad del Pais Vasco, Apartado 1072, San Sebastian 20080, Spain Institut fiir Experimentalphysik, Johannes Kepler Uni Uersita't, A-4040 Linz, Austria Laboratoire des Collisions Atomiques, Universite de Bordeaux I, 351 Cours de la Liberation, 33405 Talence CEDEX, France Centro Atomico Bariloche and Instituto Balseiro, 8400 Bariloche, Argentina (Received 16 March l 992) The energy loss of H projectiles in solid and gaseous Zn has been measured for the first time down to 20 keV/u, showing large differences in the stopping cross section depending on the state of aggregation of the target. For 25-keV protons, the stopping cross section of Zn in the gas phase is found to be 60% higher than that in the solid phase. A charge-state approach to the stopping power of ions in the solid and in the gas is successful in explaining this effect. PACS numbers: 79.20.Nc, 34.50. Bw, 61. 80.Mk The electronic stopping power of matter for swift pro- tons with energies below, and of the order of, 100 keV is dominated by outer-shell electron excitation and ioniza- tion. The stopping power has its maximum in this energy region for almost every target. As a result of the different electronic structure of the outer shells of a ma- terial in different states of aggregation, especially in the case of metals, one would expect large differences be- tween solid and gas phase stopping for the same sub- stance. Up to now, the few reported measurements either deal with materials for which the electronic structure does not change appreciably between the two phases [1, 2], or are just measurements of the ratios of the He to H stopping cross sections [3]. These data do not permit the testing of theoretical predictions since they are re- stricted to simple elements [4-6]. In the low-energy range (below 100 keV for protons) accurate first- principles calculations are only possible for light elements like H and He targets [7,8]. From the experimental side, it is not easy to find an element for which the two phases show a marked difference in the valence structure in a temperature range where the experiment is feasible. Compounds would not be suitable because they inevitably add other aggregation effects due to the chemical bonds. We have chosen Zn as a target as it has a relatively low boiling point and is easy to handle in the solid phase. The experiment was performed at the University of Linz using the 700-keV Van de Graaff accelerator. The energy loss in the gas phase was measured by transmis- sion of the ion beam through a vapor cell in a way similar to our earlier measurements on water vapor [2]. Protons and deuterons in the energy range E;„=15 keV/u to E~, „=720 keV/u were used as projectiles. After transmission through the gas cell, some of the ions im- pinged on a scattering target (a thin platinum layer evap- orated onto carbon). The energy of the scattered ions was detected at an angle of 90 with respect to the beam axis by means of a particle-implanted silicon detector. The detector and the whole amplification system were connected to a thermostat in order to minimize thermal drifts. The vapor cell had a length of 30 cm with small apertures of 1. 5 mm diameter for the ions to enter and leave. In the cell, the zinc vapor was in thermal equilibri- um with the condensed phase (usually the liquid). Be- cause of the low dissociation energy of Zn clusters the va- por consisted essentially of Zn atoms. The temperature of the gas cell was chosen and kept constant via heating, which was applied at the outer surface of the cell. A homogeneous temperature profile was obtained. The temperature was controlled by a thermocouple to keep the vapor pressure constant. Reproducibility was possible within ~5%. Hence we measured the energy losses, BE(E, n), at a constant ion energy E for a set of vapor densities n in the range 8&& IO' to 6&10' atoms/cm . The slope of the linear regression of hE (E, n) vs hE (E, „, n) at constant energy E yielded the ratio of the stopping cross sections S (E)/S (E, „). By fixing S(E, „) equal to the theoretically calculated value at 700 keV (see below) we obtain the stopping cross section at all energies. In order to measure the charge-state fractions of pro- tons and neutral hydrogen atoms, p+ and po, in the ion beam after exiting the vapor cell, a magnetic deflection field was applied, preventing the charged projectiles from hitting the scattering target. The scattered intensities Io and It,t with and without applied magnetic field, respec- tively, were measured during equal times (100 s) under stable current conditions. This yielded p =lo/It, t, and p+ =1 — P (P was found to be negligible). The charge-state fractions obtained follow a common curve with uncertainties of at most 2%. The stopping cross section (SCS) for the solid phase was determined by the well established Rutherford back- scattering (RBS) technique [9, 10]: Zinc was evaporated onto a carbon backing following Ref. [11];purity, stabili- 1992 The American Physical Society 1137