Theoretical study on the effect of an intermediate layer in CIS-based ETA-solar cells
Catelijne Grasso*, Marc Burgelman
Ghent University, Electronics and Information Systems (ELIS), Pietersnieuwstraat 41, B-9000 Ghent, Belgium
Abstract A theoretical treatment is given of the mechanisms that might be responsible for reported improvements in solarcell characteristics of, in particular, TiO2 yCuInS2 based nanostructured ETA-solar cells. Especially, the possible benefits of the insertion of an intermediate tunnel barrier (Al2 O3 , MgO) or buffer layer (CdS) are investigated. Three mechanisms to improve Voc without compromising Jsc are discussed: (i) introduction of an energy barrier which is higher for dark current than for light current;(ii) reduction of the interface state density at either side of the intermediate layer (IL) compared to the structure without IL, for materials reasons; and (iii) reduction of the effect of interface states due to a more favourable position of the Fermi levels at the interface, with respect to the band edges. The dark saturation current and hence Voc turn out to be determined by interfacerecombination, and chemical and electrostatic interactions at the interfaces. 2003 Elsevier B.V. All rights reserved.
Keywords: Solar cell; ETA; CuInS2; Tunneling
1. Introduction Insertion of very thin insulating oxide coatings (MgO, Al2O3, ZnO, TiO2, ZrO2, Y2O3) at the active interface of nanostructured solar cells has led to significant improvement of solar cell characteristics. This holds both fordye-sensitized cells w1–3x and all solid state ETA-solar cells w4x. These ETA-solar cells are designed according to a recently introduced solar cell concept, applying an Extremely Thin inorganic Absorber layer such as CuInS2 w5x or CdTe w6x at the interface of a transparent nano- or microstructured p–n heterojunction, e.g. n-TiO2 yp-CuSCN. Recently also improvement of cell characteristics on applyingan intermediate noninsulating buffer layer (CdS, In2S3) in a TiO2 yCuInS2 ETA-system is reported w7x. Suggestions for the physical reasons for the observed cell improvement are given, but a quantitative understanding and an insight into the limits of cell improvement to be expected are still largely lacking. In other cell systems though, the action of an intermediate layer (IL) is well studied;we mention a wide gap insulating layer used in metal-insulator-semiconductor cells w8,9x
*Corresponding author. Tel.: q32-9-264-8953; fax: q32-9-2648961. E-mail address: email@example.com (C. Grasso). 0040-6090/04/$ - see front matter doi:10.1016/j.tsf.2003.11.008
and a conducting buffer layer (CdS, In2S3,«) in planar CIGS solar cells. In this article, the theoretical base isstudied for possible improvements in cell current and voltage for TiO2 yCuInS2 based nanostructured solar cells. In particular, we consider the potential benefits that can be expected on the insertion of an IL at the active interface. The IL can be an insulating tunnel barrier oxide, e.g. Al2O3 or MgO, or a conducting buffer layer of CdS. Interfaces are suspected to play a significant role in this,since in nanostructured cells, materials interfaces are never further away than ;25 nm, and the microscopic surface and interface area is very large. This means that especially interface recombination (IFR) is on suspicion. 2. Theory In (solid state) dye-sensitized solar cells (DSSC), the light current JL is more or less acceptable but the open circuit voltage Voc is too low; for ETA-solar cells eventhe light current is not very good. According to the equation Vocs kT B JL E lnC F Žfor JL4Js., q D Js G (1)
a low value of Voc can be ascribed to a dark saturation
2003 Elsevier B.V. All rights reserved.
C. Grasso, M. Burgelman / Thin Solid Films 451 – 452 (2004) 156–159
Fig. 1. (inset) Tunneling through an energy barrier (f1 , f2) of width s. (main) J–V characteristics of...