Susan Schorr, Christiane Stephan and Christian A. Kaufmann

Abstract In order to understand the importance of the structural properties of compound semiconductors for the operation of a thin-film solar cell, this section aims to explain the operation principle using the example of a Cu(In, Ga)Se2 (CIGSe) thin-film solar cell. For detailed information the reader is kindly referred to the literature for a more extensive overview of the recent developments [1], device operation [2] and material preparation [3].

5.1 Introduction

CIGSe thin-film solar cells are made of a stack of metal and semiconductor thin films in the following sequence: a molybdenum back contact (metal), a polycrys­talline CIGSe absorber layer (p-type semiconductor), CdS buffer layer (n-type semiconductor), and ZnO front contact (n-type semiconductor). Together, the CdS and the ZnO are often referred to as the ‘window’ of the device. The core of the device is the p-n-heterojunction between the p-type absorber layer and the n-type window layers. The resulting energy band line-up and a cross sectional scanning electron microscope view of a complete device are shown in Fig. 5.1. Due to the high doping of the ZnO front contact layer the field, which develops upon contact of the n — and p-type materials in the interface region, is located almost entirely inside the absorber layer. In comparison to homojunctions, the heterojunction has the advantage that the n-type component can be chosen such that its band gap is

S. Schorr (H) • C. Stephan

Institut Fur Geologische Wissenschaften, Freie Universitat Berlin,

Malteserstr. 100, 12249 Berlin, Germany e-mail: susan. schorr@fu-berlin. de

C. Stephan

e-mail: christiane. stephan@fu-berlin. de S. Schorr • C. A. Kaufmann

Helmholtz-Zentrum Berlin Fur Materialien Und Energie, Berlin, Germany

© Springer International Publishing Switzerland 2015 83

G. J. Kearley and V. K. Peterson (eds.), Neutron Applications in Materials for Energy, Neutron Scattering Applications and Techniques,

DOI 10.1007/978-3-319-06656-1_5

large and the photoactive band-gap Eg of the solar-cell device, determined by the band gap of the absorber-layer material, is optimized to reach high conversion — efficiencies.

Figure 5.2 illustrates the basic working principle of a CIGSe thin-film solar cell. Electron-hole pairs are generated by light absorption within the absorber thin-film. Absorption of photons with an energy higher than Eg results in the loss of excess energy via thermalization. If an electron-hole pair is excited within the depletion

region near the font interface of the device, it is immediately spatially separated by the electric field that is present. If generation happens outside the depletion region, the minority charge carrier has to diffuse into the depletion region in order to contribute to the photo current. Recombination of the electron-hole pairs can occur by a number of processes: radiative and non-radiative, band-to-band or via one or several defect levels located in the band gap of the absorber material, within the bulk of the thin-film material, within the depletion region, and at the front interface of the device or possibly at the grain boundaries of the polycrystalline CIGSe thin — film. The nature of grain boundaries in CIGSe thin-films however, has proven to be considerably more benign than in other semiconductor materials such as for example silicon, the exact reasons for which are still under discussion [4].

CIGSe absorbers for photovoltaic application are mostly fabricated slightly Cu-poor and with an overall Ga content of [Ga]/([Ga] + [In]) * 0.3. The resulting CIGSe material is ap-type semiconductor material (Eg * 1.15 eV), which is highly compensated. This means that there are acceptor — and donor-type defects present within the material and p-type conductivity is established due to the exceptionally-low formation energy of Cu vacancies. In addition, the formation of defect complexes, such as 2VCu + InC+,, seems to play an important role in terms of the electronic properties of the material, and also regarding phase formation and crystal structure [5] (Figs. 5.2 and 5.3).

For most applications rigid soda-lime glass is used as a substrate material, but flexible metal or polyimide foils have been used successfully. Working on soda — lime glass, it has been established that Na plays an important role in increasing the carrier concentration within the CIGSe absorber [6]. It diffuses at elevated process temperatures from the glass substrate through the Mo back contact into the growing CIGSe thin-film [7], and also has an effect on the morphology and material inter­diffusion in the growing layer [8, 9]. On samples, which do not intrinsically contain sodium, it has to be externally supplied.

Fig. 5.3 Elemental flux and substrate-temperature profiles for a typical 3-stage co-evaporation process

While ZnO and Mo are usually sputtered, CdS is most widely deposited via chemical-bath deposition. As the use of Cd-containing components is viewed critically, Cd-free alternatives are in development and, in some cases, already incorporated in the commercially-available product [10]. The methods for CIGSe thin-film fabrication can be grouped into two main categories: co-evaporation and sequential processing [10]. The method of co-evaporation relies on the more or less simultaneous evaporation of the elements Cu, In, Ga, and Se to form a CIGSe thin — film on a heated, Mo-coated substrate in vacuum. For sequential processing on the other hand, a so-called precursor layer in either metallic, binary, or nanoparticle form is treated in a reactive atmosphere in order to make the CIGSe thin-film absorber. The current record conversion-efficiency of 20.3 % [11] is reached by a CIGSe thin-film device that was deposited via 3-stage co-evaporation [12].

The 3-stage co-evaporation process for CIGSe deposition leads to a thin film, which has a characteristic compositional in-depth profile, shown in Fig. 5.4. While Ga accumulates near the Mo/CIGSe interface during the inter-diffusion of Cu-Se in stage 2, the last stage of the process also provides for a slightly increased Ga content near the surface. Of course the latter depends on the Ga flux during stage 3. The Ga profile is of relevance to the resulting solar-cell device, as the Ga content determines the energy band-gap of the CIGSe material. A graded compositional profile is therefore equivalent with an in-depth band-gap grading within the device. Compositional gradients, as shown in Fig. 5.4, can be observed even within single grains. Not only does the Ga content have an impact on the resulting energy band — gap of CIGSe, but also on the Cu-deficiency, in particular near the absorber surface. This is possibly caused by the nature of the last stage of the 3-stage process, which can affect its band gap. Cu-deficient CIGSe phases, such as Cu(In, Ga)3Se5, show slightly-larger band gaps than the stoichiometric compound. The presence of such Cu-poor phases at the front interface has been argued to be of relevance for efficient CIGSe thin-film devices [13]. Other evidence points towards entirely Cu-free film surfaces, which ensure favourable interface formation between the CIGSe and the buffer layer [14]. It is most likely that the exact process-parameters determine which

of the different scenarios holds in each case. CIGSe thin-films from other manu­facturing routines, such as the sequential processing, will have grown under very different thermodynamic and kinetic growth conditions and, so far, only little is known of the extent to which the resulting thin films can be considered identical.

In order to understand the correlation between growth conditions, material properties and final device quality, much has to be understood regarding the basic material properties of the absorber material in question. Topics such as the occu­pation density of the different sites within the crystal lattice, i. e. defect formation under certain growth conditions, at certain material compositions, or in the presence of foreign elements are areas where neutron diffraction can provide valuable input for the design of growth models, analytical material science, and also for compu­tational methods.