Impact of 600 eV protons at grazing incidence on LiF(100) is studied with a new coincidence technique combining energy loss and electron spectroscopy. Correlation between the secondary electrons and the charge state of the scattered projectiles demonstrates the role of the H 2 ions formed on the surface as precursors for electron emission. However, the main channel for energy loss is not associated with electron emission but is interpreted as the population of surface excitons. PACS numbers: 79.20.Rf, 34.50.Dy, 34.70. + e, Since the pioneering work of Souda et al.[1] ten years ago, the study of low-energy ions interacting with large band gap insulators and the subsequent energy loss and electronic emission has attracted much interest [2][3][4][5][6][7]. Compared to metallic surfaces, the large band gap and high binding energies of the valence electrons are expected to induce profound differences from several aspects: (i) The energy loss (through electronic stopping) of ions traveling along the surface should exhibit a threshold behavior with incident energy. (ii) The ion velocity threshold for kinetic electron emission is expected to increase with respect to that of metals. (iii) The resonant electron transfer (from and to the solid) should be strongly reduced. Points (i) and (ii) are simply due to the much larger energy required to excite or ionize valence electrons. The most recent observations by Auth et al. [8] have confirmed point (i) in collisions of protons with LiF but with a threshold behavior appearing only at substantially lower energy than expected [9]. With respect to point (ii), Vana et al. [2] showed that no clear energy threshold can be observed in the secondary electron emission yield during singly charged ions (H 1 , Ar 1 ) collision on LiF, whereas a threshold of 1 keV was measured for the same projectiles on Au. As for charge exchange, the resonant neutralization of singly charged alkali ions and the resonant ionization of alkali atoms is strongly reduced because of the band gap [10]. The suppression of this electron loss channel is also partly responsible for the surprisingly large negative ion fractions, up to 60%-90% [11,12] observed for oxygen or fluorine interacting with alkali halides. The capture process was elucidated only recently as being due to a lowering of the projectile affinity level in the Madelung potential [8,13,14].The presence of a large band gap indeed controls the resonant electron capture and loss but does not seem to play the same decisive role in the energy loss or in the electron emission. This paradox has been studied in detail for the H 1 -LiF system, which may be considered as a reference. The projectile has a well-known electronic structure, and since the LiF band gap extends above the vacuum level, energy loss and electron emission should be intimately related. These studies have called for the existence of an intermediate state able to reduce the effective band gap during the collision process. On one hand, independent measurements of secondary electron yields [...

No abstract

No abstract