ELSEVIER Surface Science 364 (1996) 242-252 surface science The structure of the Ge(001)-(2 x 1) reconstruction investigated with X-ray diffraction X. Torrelles, H.A. van der Vegt ~, V.H. Etgens 2, P. Fajardo, J. Alvarez, S. Ferrer * European Synchrotron Radiation Facility, BP 220, 38043 Grenoble Cedex, France Received 18 October 1995; accepted for publication 23 February 1996 Abstract The atomic structure of the room temperature (2 × 1) reconstruction of Ge(001) has been investigated with X-ray diffraction measurements by collecting an extensive and accurate set of diffracted intensities. The structural model that best agrees with the data consists of a buckled array of disordered dimers: there is a 0.5 probability of finding one of the two dimer orientations (positive and negative tilt angles) in any unit cell. The dimer tilt angle is found to be 15.6+0.6 °, its bond length is expanded by 4% compared to the bulk bond length, the reconstruction is found to extend to eight atomic layers and the maximum distortion of the bonds has been found to be less than 6%. The model is compared with similar models, and the dynamical and ordered models and the dynamics of the dimer flipping are discussed. Keywords: Germanium; X-ray diffraction 1. Introduction The atomic structure of the clean (001) surfaces of Si and Ge has been studied with most, if not all, the appropriate surface techniques. The basic (2 x 1) reconstruction is generally accepted to entail the formation of dimers, created through pairing of nearest-neighbour surface atoms 1,1], having their bond axis slightly inclined (10-20 °) with respect to the surface plane (buckled dimers) * Corresponding author. E-mail: ferrer@esrf.fr. 1 Permanent address: FOM Institute for Atomic and Molecular Physics, Kruislaan 407, 1098 SJ Amsterdam, The Netherlands. 2 Permanent address: CNRS-Laboratoire de Minera- logie-Cristallographie, Universit~ Paris VI, 4 Place Jussieu, 75005 Paris Cedex, France. 1,2-4]. The only apparent discrepancy to this large number of experimental findings came from STM that showed symmetrical-looking (i.e. non-buck- led) dimers at room temperature [5]. This appar- ent contradiction was solved by realizing that the dimers would probably be flipping between two symmetrical buckled configurations (i.e. from say, atom 1 up, atom 2 down to atom 1 down, atom 2 up) and that the scanning microscope was collect- ing an averaged electron density. This picture was supported by the theoretical work of Dabrowski et al. I-6], which showed that the two buckled configurations may be described by a double potential energy well with an activation energy of 100 meV separating both stable configurations. Based on these ideas, a detailed X-ray surface diffraction study on Ge was undertaken by Rossman et al. [8] to investigate the buckled nature and the dynamics of the dimers. These 0039-6028/96/$15.00 Copyright © 1996 Elsevier Science B.V. All rights reserved PII S0039-6028 (96) 00650-4