Review: Progress in solar cells from hydrogenated amorphous silicon

Abstract

Hydrogenated amorphous silicon (a-Si:H) has been used for decades—doped and as intrinsic absorber layers—in thin-film silicon solar cells. Whereas their effiency was improved for a long time by the deposition of higher quality absorber layers, recent improvements can be attributed to a better understanding of the interfaces, allowing for their specific engineering. In this review, we briefly resume the state-of-the-art of a-Si:H solar cell technology from growth and characterization of single layers to full solar cells and multijunction devices. Focusing on the absorber layer quality first, we highlight thereafter aspects of interface problematics and discuss the growth and role of doped microcrystalline silicon-oxide layers and approaches of 3D-solar-cell designs in more detail.

Although the findings summarized in this review were obtained from thin-film solar cells, we show that a-Si:H is a very versatile material with properties that are of high interest for application in other devices such as heterojunction solar cells, detectors, or optoelectronic devices.

Introduction

Thin-film silicon (TF-Si) solar cells are one possible answer to the increasing energy demand of today. Hydrogenated amorphous silicon (a-Si:H) has played a crucial role therein—for decades already as intrinsic absorber layers with doped layers to build PIN junctions, and to an increasingly important extent in combination with crystalline silicon wafers in heterojunction (HIT) solar cells [1]. This article is devoted to single-junction solar cells with a-Si:H absorber layers; however, many findings are relevant to other a-Si:H applications too.

Focusing on device performance, an in-depth discussion of general properties of a-Si:H is beyond the scope of this review and we only introduce the basic concepts that are necessary for the further understanding; the interested reader is referred to the large amount of literature discussed in [2], [3].

Fig. 1 shows the performance of laboratory-scale record solar cells as a function of the bandgap for 1000 W/m² irradiation with spectrum AM1.5g [4]. The values of open-circuit voltage (V_oc), fill factor (FF), short-circuit current density (J_sc), and solar cell conversion efficiency of world-record solar cells are from [5], indicated by markers. The bandgap values are from [6] (crystalline Si, CdTe, GaAs, InP), [7] (perovskite), [8] (CIGS), and from our own measurements of a-Si:H. The lines indicate empirical trends (V_oc = E_g/q and V_oc =2/3) and physical limits of these parameters. The calculations are compared in appendix A of [9] for the models denoted single-pair [9], Shockley-Queisser [10], Kiess [11], and Green [12], [13], [14].

The performance of a-Si:H solar cells is well below fundamental conversion efficiency limits. Although J_sc and V_oc have potential of improvement too, the FF contributes most to the difference between device performance and potential. While high-efficiency a-Si:H solar cells have a FF above 75% and an efficiency above 11% as deposited, the FF and consequently the efficiency drop by typically 10–20% within the first few months of operation. This light-induced degradation (LID) is characteristic for a-Si:H cells and will be discussed at several places in this manuscript. On the module level, LID leads to a significant initial efficiency drop that is accounted for in advertised efficiencies; on long-term, the degradation of a-Si:H solar modules is linear and with about 1%/yr close to other solar technologies [15].

After an overview over the historic development, configurations, and applications of solar cells with an a-Si:H absorber layer in Section 2, we will detail the synthesis of a-Si:H by different methods in Section 3. After an introduction to the specific characteristics of the bandgap of a-Si:H, Section 4 discusses recent improvements in the understanding of the nature and creation of defects in a-Si:H. Section 5 highlights advances in characterization techniques of both single layers and fully operational devices that are particularly useful for a-Si:H solar cells. In Section 6, we present results correlating deposition parameters with absorber layer and solar cell properties. Gathering the puzzle pieces of these subjects together, a consistent picture of the performance of a-Si:H evolves, showing its potential, but also limitations with emphasis on LID.

Whereas LID is to some extent related to the amorphous nature of a-Si:H, other limitations of the a-Si:H solar cell performance are largely caused by interfaces. In Section 7, we will present examples of approaches to overcome both limitations of bulk and interfaces. Finally, we will conclude the manuscript and give an outlook to future applications of a-Si:H solar cells in Section 8.

Part of this manuscript was published in modified form in the PhD theses of M. Stuckelberger [9] and R. Biron [16].