No abstract

We analyze the dependence of the interface defect density Dit in amorphous/crystalline silicon (a-Si:H/c-Si) heterojunctions on the microscopic properties of ultrathin (10 nm) undoped a-Si:H passivation layers. It is shown that the hydrogen bonding and network disorder, probed by infrared- and photoelectron spectroscopy, govern the initial Dit and its behavior upon a short thermal treatment at 200 °C. While the initial Dit is determined by the local and nonequilibrated interface structure, the annealed Dit is defined by the bulk a-Si:H network strain. Thus it appears that the equilibrated a-Si:H/c-Si interface does not possess unique electronic properties but is governed by the a-Si:H bulk defects.

Single grain boundaries in CuGaSe 2 have been grown epitaxially. Hall measurements indicate a barrier of 30 -40 meV to majority carrier transport. Nevertheless, local surface potential measurements show the absence of space charge around the grain boundary; i.e., it is neutral. Theoretical calculations [Persson and Zunger, Phys. Rev. Lett. 91, 266401 (2003)] have predicted a neutral barrier for the present 3 grain boundary. Thus, we have experimentally shown the existence of a neutral grain-boundary barrier, however, smaller than theoretically predicted. DOI: 10.1103/PhysRevLett.97.146601 PACS numbers: 72.80.Ey, 61.72.Mm, 71.55.Gs, 72.20.ÿi The electronic structure of Cu chalcopyrites shows numerous peculiarities, including, i.e., their grain-boundary properties. The first model [1,2] for the electronic structure of grain boundaries in chalcopyrites was based on the Seto model [3], where charged defects at grain boundaries cause a barrier for majority carriers (holes in the chalcopyrites discussed here). Barriers and space charge regions have been found in polycrystalline films by Hall measurements and Kelvin probe force microscopy (KPFM) (see, e.g., [1,4 -6] ). Theoretically it has been predicted that grain boundaries in chalcopyrites represent a barrier without charged defects [7,8]. In the Cu chalcopyrites, the top of the valence band is formed from antibonding Cu d states, which lifts it to higher energy compared to the corresponding II-VI compounds [9]. Grain boundaries consisting of f112g tet planes in the tetragonal system [10], corresponding to the f111g cub planes in the cubic system, are proposed to be Cu deficient, which results in a lowering of the valence band maximum. This results in a barrier in the valence band without a space charge at the grain boundary. Recently it has been argued that only a specific structure of a f112g tet grain boundary leads to this barrier and that the experimentally observed structures should not show a neutral barrier [11].Grain boundaries along the f112g tet planes can be described as twins. The coincidence site lattice (CSL) of such grain boundaries is characterized by a value of 3, the lowest value possible. Thus, it is expected that grain boundaries along f112g tet planes show a low defect density. A generating function of the CSL for the cubic system has been derived in Ref. [12]. Since the tetragonal distortion of chalcopyrites is small and since the CSL concerns only the two-dimensional plane of the grain boundary we assume the generating function for the cubic system as an approximation for the chalcopyrite system. Then it becomes clear that a polycrystalline film with f220g tet (i.e., f110g cub ) texture and vertical grain boundaries contains predominantly 3 grain boundaries. In fact, it has been shown that polycrystalline absorber films with f220g tet texture result in higher efficiencies of the corresponding solar cells compared to the usual f112g tet texture [13]. Therefore, 3 grain boundaries are the ones appearing in the most successful solar cells ...

The lack of an efficiency increase with increasing Ga content in Cu(In,Ga)Se 2 solar cells has attracted much scientific interest. It has been claimed that the physical properties of grain boundaries are responsible for this curious effect. Here, we present an in-depth analysis of electronic potential barriers at grain boundaries (GBs) in a series of Cu(In,Ga)Se 2 (CIGSe) thin films using Kelvin probe force microscopy (KPFM) measurements, extending our previous study [Sol. Energy Mat. Sol. Cells 103, 86 (2012)]. Here, (i) we show, by comparison with data of the crystal lattice orientations, that localization of GBs purely from KPFM topography data allows reliable localization of GBs. (ii) We consider the averaging effect of KPFM due to long-range electrostatic forces for the analysis of the electronic GB properties to determine the real potential barrier height for each individual GB; we determine potential variations ranging from −400 to +400 mV. (iii) We consider the different physical origin of positive and negative potential barriers and present a quantitative analysis of the results to determine charge carrier concentration and defect densities at GBs. From our data and anaylsis we do not observe any systematic variation of these quantities with the Ga content.

Long-range electrostatic forces govern the imaging mechanism in electrostatic force microscopy as well as in Kelvin probe force microscopy. To improve the analysis of such images, simulations of the electrostatic field distribution have been performed in the past using a flat surface and a cone-shaped tip. However, the electrostatic field distribution between a tip and a sample depends strongly on the surface topography, which has been neglected in previous studies. It is therefore of general importance to study the influence of sample topography features on Kelvin probe force microscopy images, which we address here by performing finite element simulations. We show how the surface potential measurement is influenced by surface steps and surface grooves, considering potential variations in the form of a potential peak and a potential step. The influence of the topography on the measurement of the surface potential is found to be rather small compared to a typical experimental resolution. Surprisingly, in the case of a coinciding topography and potential step an improvement of the potential profile due to the inclusion of the topography is observed. Finally, based on the obtained results, suggestions for the realization of KPFM measurement are given.