E*PCOS06 Recent Progress in Understanding PC-Materials for Optical and Electronic Storage Christoph Steimer , Wojciech Welnic , Ralf Detemple, Daniel Wamwangi, Henning Dieker and Matthias Wuttig I. Institute of Physics (IA), RWTH-Aachen University, Sommerfeldstr. 14-18, 52056 Aachen, Germany ABSTRACT At present, phase change media that can meet the increasing demands of modern storage technology are developed in industrial and university labs employing experimental optimisation schemes an. Merely by try and error, for a large number of samples it could be shown that only materials with cubic or near-cubic structures show the required optical contrast for phase change recording as opposed to materials based upon tetrahedral coordination. Here we present an alternate approach, which is based upon both combinatorial techniques to produce sample libraries and ab initio calculations of the ground state and electronic structure. The calculations explain the variation of the ground state structure and the bonding type with stochiometry Understanding the observed trends and combining them with an understanding of the kinetics of the phase transitions promises the directed optimisation of future phase change media. Key words: phase change storage, density functional theory, stoichiometry, crystalline and amorphous structure 1. INTRODUCTION Phase change (PC) materials have been in commercial use in rewritable optical storage (DVD-RW e.g.) for a decade and are currently investigated as nonvolatile electronic storage to replace conventional FLASH-memory: A ps- duration laser or current pulse of high intensity is used to melt a sub-micron sized spot of crystalline material and quench it to the amorphous state. A second pulse of lower intensity but longer duration leads to a less pronounced spatial temperature profile and lower cooling rates but provides suffcient thermal energy to activate the crystallisation process. Since reflectivity and conductivity of the amorphous state are lower than the crystalline values, a third laser - or current pulse, even lower in intensity and thus not changing the material, can be used to read out the current state of the bit [13]. Surprisingly, despite their commercial application material development of PC-media still heavily relies on empirical approaches. This contribution summarizes how stoichiometry determines the structure of the crystalline and the amorphous phase. 2. STOICHIOMETRY & STRUCTURE Based on their different compositions and their different crystallisation behaviour common PC-materials can be devided into a Sb-rich and Te-rich class, which are based on Sb 2 Te or Sb 2 Te 3 and are growth -or nucleation dominated [12,13]. Two of the phases formed upon annealing of AgIn doped Sb 2 Te are AgInTe 2 and AgSbTe 2 , which were chosen as model-systems [4]. After molecular beam deposition on glass-substrates, the samples were crystallised by annealing in a N 2 -atmosphere. A static tester was used to laser amorphise a spot of a diameter of about 0.3 µm to start out from identical conditions. The reflectivity of the amorphous spot was measured before exposure to a crystallisation pulse of defined intensity and duration and a second measurement of the reflectivity with the same laser. The procedure was repeated for an array of spots with varying intensities and durations of the crystallisation pulse. The results are visualized in figure 2: While AgSbTe 2 shows an increase in reflectivity due to crystallisation expected for low intensities and long pulse durations, the low intensity wings for low power or short duration are followed by immediate ablation in the case of AgInTe 2 . To understand the non-observable change in reflectivity, the sheet-resistance of as deposited samples was monitored upon annealing in a N 2 –atmosphere.