تاثیر توزیع دم باند روی به دام افتادن حامل ها در سیلیکون بی نظم هیدروژنه برای کاربردهای سلول خورشیدی Impact of band tail distribution on carrier trapping in hydrogenated amorphous silicon for solar cell applications

۱٫ Introduction Carrier trapping is a key factor that restricts carrier transport in amorphous semiconductor devices [1–۵]. Once carriers are trapped in those devices, particularly in the active layer, the trapped carriers result in the Shockley–Read–Hall (SRH) recombination [6], and thereby the carrier lifetime is reduced. Besides, an accumulation of trapped carriers induces the quasi-Fermi level shift accompanying band bending [7,8], which may degrade the carrier collection. The carrier trapping thus impacts on the carrier recombination and collection so that studies on carrier trapping are beneficial for understanding the carrier transport and improving the device performance. In amorphous semiconductors such as hydrogenated amorphous silicon (a-Si:H), carriers can be trapped at gap states (also called localized states) associated with various defects. The gap states are usually classified into two groups: the band tail states and the mid-gap states [9,10]. The band tail states, related to the network disorder, are extended from the conduction and valence band (CB and VB) edges. The CB tail states exhibit electron traps whereas the VB tail states exhibit hole traps [6]. Because of the existence of these traps, the carrier mobility, μ, is strongly limited. On the other hand, the mid-gap states, originating from dangling bonds (DBs), are formed near the middle of bandgap. They behave as recombination centers for carriers. So, the carrier lifetime, τ, is often governed by the density of these mid-gap states [6]. Among amorphous semiconductors, a-Si:H is widely used for various device applications. For example, in thin-film Si solar cells [11,12], an a-Si:H film is used as a photovoltaic layer that plays important roles in carrier transport as well as light absorption. In amorphous/crystalline Si (c-Si) hetero-junction solar cells [13,14], a thin layer of a-Si:H is employed to passivate the c-Si surface and also selectively transfer carriers, i.e., either electrons or holes to electrodes. The passivation and selective transfer are known to be required for this kind of highefficiency solar cells [15]. So far, carrier trapping in an a-Si:H film has been well studied by time of flight technique [2–۵]. The gap states, i.e., the origin of carrier trapping, have been characterized by several methods such as constant photocurrent method (CPM) [16–۱۹], Fourier transform photocurrent spectroscopy (FTPS) [20,21], modulated photocurrent (MPC) spectroscopy [22], and deep level transient spectroscopy (DLTS) [23–۲۵]. Particularly, the density of mid-gap states, related to DBs, have been quantified by electron spin resonance (ESR) [26,27]. Furthermore, the distribution of the mid-gap states is studied by dual-beam photoconductivity (DBP), in which the bias light is used to precisely control the quasi-Fermi level [28]. Nevertheless, the impact of the gap state distribution on carrier trapping and the device performance have not been investigated systematically. In this paper, we investigate carrier trapping in intrinsic a-Si:H films from the viewpoint of the band tail distribution. The trapped carrier (electron) density at the CB tail are determined quantitatively, using an optical pump-probe technique. This technique has been already applied to in-situ monitoring of film growth processes in a-Si:H [29]. Because of its high sensitivity and convenience, we extend this technique to ex-situ characterization of a-Si:H in this article. One advantage of this technique is that one can evaluate the density of electron traps in the samples with a thickness of up to several hundred nm [29], which includes the typical thickness in the state-of-art a-Si:H solar cells [11,12]. This is in contrast with the case of time-of-flight technique, where the samples having a thickness of several micron meters are necessary [2–۵]. Using this pump-probe technique and CPM, a correlation between the trapped electron density and the VB tail broadening is studied. The effects of carrier trapping on the device performances are examined in a-Si:H p-i-n solar cells. The paper consists of the following parts. In the next section (Section 2), a theory for carrier generation and recombination is briefly described to derive the trapped carrier density. In experimental section (Section 3), we explain sample preparation, trap characterization techniques, and a-Si:H solar cell structure. In results and discussion (Section 4), the trapped carrier density, the tail state distribution and the device performance are reported. The occupation, the origin of traps, carrier transport and trapping are then discussed. Finally, the solar cell performances are examined in terms of carrier transport and trapping.