Abstract. Uppaal is a tool suite for automatic verification of safety and bounded liveness properties of real-time systems modeled as networks of timed automata. It includes: a graphical interface that supports graphical and textual representations of networks of timed automata, and automatic transformation from graphical representations to textual format, a compiler that transforms a certain class of linear hybrid systems to networks of timed automata, and a model-checker which is implemented based on constraint-solving techniques. Uppaal also supports diagnostic model-checking providing diagnostic information in case verification of a particular real-time systems fails. The current version of Uppaal is available on the World Wide Web via the Uppaal home page

No abstract

No abstract

Tandem solar cell structures require a high-performance wide band gap absorber as top cell. A possible candidate is CuGaSe 2 , with a fundamental band gap of 1.7 eV. However, a significant open-circuit voltage deficit is often reported for wide band gap chalcopyrite solar cells like CuGaSe 2 . In this paper, we show that the open-circuit voltage can be drastically improved in wide band gap p-Cu(In,Ga)Se 2 and p-CuGaSe 2 devices by improving the conduction band alignment to the n-type buffer layer. This is accomplished by using Zn 1−x Sn x O y , grown by atomic layer deposition, as a buffer layer. In this case, the conduction band level can be adapted to an almost perfect fit to the wide band gap Cu(In,Ga) It has been proven difficult to maintain a good device quality when the gallium content is increased. The main problem is that the opencircuit voltage (V oc ) does generally not scale with the band gap energy as predicted, and it tends to saturate for absorber band gap energies above roughly 1.3 eV. 7,8 This lack of performance in high gallium CIGS devices has been the subject of numerous studies. 7-11The dominating recombination paths, in CuGaSe 2 devices, have been assigned to tunneling enhanced recombination either in the space-charge region or at the interface. 12,13 A possible explanation is trap states formed by cation anti-site (In Cu /Ga Cu ) or anion vacancy (V Se ) defects that become deeper positioned within the band gap when the gallium content increases, and thereby form more effective recombination centers. 11 Furthermore, the difference between the Fermi level and the valence band energy at the absorber surface seems to remain constant around 0.8 eV, independently of gallium content. 13 Consequently, the Fermi level position at the absorber/buffer interface is closer to the middle of the band gap at high gallium contents, and no beneficial type inversion can be expected. Thus, the influence of recombination close to or at the interface appears to become more prominent when the CIGS band gap is widened. The bulk recombination rate is also expected to increase with a large number of these defects, but it has not In the first section of this study, the growth and material characterization of nongraded CIGS absorbers with varying gallium contents (0.3 ≤ GGI ≤ 1) deposited in a single-stage co-evaporation process are described. These absorbers are applied in the second section, where we use the temperature dependence of ALD grown ZTO buffer layers to show that the V oc deficit in wide band gap CIGS can be reduced by improving the absorber/buffer conduction band alignment.In the last section, we further investigate the effect of an improved band alignment by using CuGaSe 2 absorbers of higher material quality, evaporated in a 3-stage type process. 21-24It can be observed in the SEM images that the grain size is reduced with increased GGI. This general observation is often reported in the literature. The grain size reduction has been found to be influenced by a number of factors, such as film t...

The purpose with the present investigation has been to support and explain the experimental observation made regarding the enhancing effect by N doping on especially the diamond (100)-2 9 1 growth rate. Within the present study, also the commonly observed diamond (111) and (110) surfaces were included, all assumed to be H-terminated. Density functional theory calculations were used, based on a plane wave approach under periodic boundary conditions. It was shown that the surface H abstraction reaction is most probably the rate-limiting step during diamond growth. In addition, the results showed that it is N, substitutionally positioned within the upper diamond surface, that will cause the growth rate improvement, and not nitrogen chemisorbed onto the growing surface in the form of either NH (or NH 2 ). The here presented numerical value for the growth rate enhancement for the diamond (100)-2 9 1 surface is almost identical with the experimentally obtained one (3.7 vs. 3.6). In addition, the (111) and (110) surfaces were shown to undergo a different growth rate enhancement, with about half as much for the (111) and (110) surfaces as compared to the diamond (100)-2 9 1 surface (1.9, 1.7 vs. 3.7). Despite the rate improvement for all surface planes, this difference will bring about a preferred diamond (100) surface texture.