Nucleation of semiconductor Quantum Dots on nanomesas : role of stressors and early stages of capping process.
- PDF / 434,244 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 39 Downloads / 161 Views
0901-Ra13-03.1
Nucleation of semiconductor Quantum Dots on nanomesas : role of stressors and early stages of capping process. Cyril Meynier and Catherine Priester IEMN/ISEN, CNRS-UMR 8520, BP 60069 F-59652, Villeneuve d'Ascq Cedex, FRANCE. ABSTRACT Here is reported, from the theoretical point of view, how a stressor layer included in a nanomesa can modify the growth process of a mismatched material : critical thickness for 2D-3D transition, and optimal location of nucleating dots by means of the induced stress field. This stress field depends on both design parameters and sign of the misfit between the stressor and the mesa material. This is exemplified by the case of InAs deposition on InP mesas which include an InGaAs stressor. A second case in which stress field plays a key role is the capping of dots on top of nanomesas . We investigate the case of Si capping of a system where one Ge dot covers the top of each mesa of a very dense array of Si nanomesas. Upon several basic assumptions, different capping processes are simulated, in order to predict the more stable one. We point out very different behaviors for pure Si capping or SiGe capping. How long it takes for recovering a flat surface strongly depends on the presence of a thin Ge surface layer. INTRODUCTION The use of an array of nanomesas has proved to be an attractive way to obtain, for adequate design parameters of the mesas, the growth of one quantum dot per mesa. This has been recently investigated, for Ge deposited on an array of Si nanomesas, from both experimental [1] and theoretical [2] points of view. The stress field present in a system is known to strongly influence the growth process. This is exemplified by two cases in which mesas are involved. The first one deals with InP mesas , obtained by classical lithography methods, of the order of tens to a few hundreds nms wide, in which is included a stressor, not too far away from the mesa top. Next section focuses on the role of this stressor depending on the sign of its misfit with InP and the mesa size. The second case will focus on mesas smaller than a few tens of nanometers which result from the selective etching of a layer twist-bonded onto a substrate. In this case, the quantum dots one gets are square base pastilles that cover the top of the mesa. We address the question of capping such an array of nanodots. ROLE OF STRESSORS When growing, on an InP substrate, InAs quantum dots not too elongated, and choosing somewhat their location, patterning of the substrate is required [3]. Mesas appear to be attractive [46] for localising dots. However, because of the elastic relaxation at mesa edges, a mismatched material deposited on a mesa is less strained than when deposited on a flat substrate, quantum dots should then nucleate later on top of the mesas than between them. The idea is therefore to include a stressor within the mesas as schematized in figure 1. The stressor can be either tensile (Ga-rich InGaAs layer) or compressive (In-rich InGaAs layer). We have calculate the stress field in such
Data Loading...
