Introduction

Hydrogenated nanocrystalline silicon (nc-Si:H) or micro-crystalline Si (^c-Si:H) films appear as a promising material for thin-film devices such as solar cells, thin-film transistors, and sensors, since the electron mobility of nc-Si:H is much improved compared to that of hydrogenated amorphous silicon films (a-Si:H). Understanding the electronic transport and optical properties of nc-Si:H is very important for improving such device properties. As schematically shown in Figure 1, structure of nc-Si:H is very complex and is characterized by "crystalline grain clusters" (region A) which consist of crystalline grains from nano-size to ~20 nm, and it is surrounded by "disorder zones" (region B) [1, 2]. Region A forms columnar clusters of much larger dimensions extending perpendicular to the substrate. Although the influences of the grain size and crystalline volume fraction on the electronic transport have been discussed, the details of their quantitative natures are still not clear [1-8].

To understand the physical properties in such complex materials, a percolation approach may be useful. A percolation path for electronic transport in nc-Si:H films has been suggested [3]. In fact, three-dimensional conductance network calculations (computer simulations) for the conductivity, a, and Hall mobility, show an existence of critical threshold of such percolation path [9]. It is suggested that an effective medium approximation (EMA) [10] is useful for explaining the quantitative transport and optical properties in inhomogeneous systems [11]. In the present chapter, current understanding of electronic and optical properties of nc-Si:H films is briefly reviewed along the idea of EMA in which random mixture of particles (region A and B) is assumed for simplicity [11].

Nonactivated behavior of electronic transport (temperature-dependent dc conductivity), on the other hand, is one of the important and unsolved problems, which can be related to complicated structure of nc-Si:H films. In some cases, the dc conductivity seems to be proportional to exp(-B/T1/2) [12, 13]. Reasons for these behaviors have been discussed in terms of hopping conductivity in a traditional manner [12, 13]. In this chapter, the non-activated behavior can be, alternatively, explained in terms of thermionic emissions of electrons over random barriers between crystalline clusters [14].

0 0

Post a comment