High-resolution photoemission of the Sn 4d core level of Sn=Ge 111 -3 3 resolves three main components in the line shape, which are assigned to each of the three Sn atoms that form the unit cell. The line shape found is in agreement with an initial state picture and supports that the two down atoms are inequivalent. In full agreement with these results, scanning tunnel microscopy images directly show that the two down atoms are at slightly different heights in most of the surface, giving rise to an inequivalentdown-atoms (3 3) structure. These results solve a long-standing controversy on the interpretation of the Sn 4d core-level line shape and the structure of Sn=Ge 111 -3 3 . DOI: 10.1103/PhysRevLett.100.026103 PACS numbers: 68.35.ÿp, 68.37.Ef, 68.47.Fg, 79.60.ÿi Materials that are characterized by a strong interplay between different degrees of freedom tend to exhibit complex physical phenomena, difficult to understand within conventional notions [1]. A conspicuous example in lowdimensional systems is the (3 3) structure formed by 0.33 monolayers (ML) of Sn atoms on a Ge(111) surface [2]. In this structure, one Sn atom per surface unit cell is at a higher level than the other two [3][4][5]. The electronic and lattice degrees of freedom are coupled, so that the atom displaced upwards receives charge from the two down atoms (one-up, two-down model, 1U2D) [4]. A delicate balance between elastic and electronic energies stabilizes this phase only in the 25-220 K temperature range. Above 220 K, thermal induced vertical fluctuations destroy the (3 3) long-range order [6], and below 25 K, a flat, Mott insulating phase is formed [7].The 1U2D model for the surface structure is supported by compelling experimental [2 -12] and theoretical [3,4,4 -16] evidence. However, a controversy lasting almost a full decade affects the interpretation of the Sn 4d core-level line shape and contradictory results are obtained from core-level related structural techniques. Early work identified two different Sn 4d components [17], which were related later on to the up and down Sn atoms. The lowest binding energy (BE) component was found to be more intense and thus was attributed to the down atoms, which are twice as many as the up atoms [4,18,19]. However, within an initial state photoemission picture, the lowest BE component is expected to originate from up Sn adatoms, which are charge acceptors. The apparent exchange was attributed to unclear site-dependent final-state screening effects [4,18]. Later on, doping experiments [20] showed the filled valence character of the lowest BE component. These two arguments support a unit cell containing two up atoms (the 2U1D model). Finally, core-level photoelectron diffraction (PED) concludes that down adatoms correspond to the lowest BE component [21], while chemically resolved x-ray standing wave experiments support the opposite assignment and find 2U1D fluctuations (at 300 K) [22]. This long-standing debate on the interpretation of the Sn 4d line shape reaches far beyond this particular ...