Highly efficient silicon solar cells require excellent surface passivation. Al2O3 is now the standard material for p-type silicon as high-quality passivation. At NaMLab, homogeneous Al2O3 layers were further developed to multi-oxide nanolaminates with tailored material properties. These layers enhance the classical surface passivation and open the field of conductive passivation for next generation PERC (Passivated Emitter and Rear Contact) cells.
(I) Classical passivation: engineering of interface charge carriers. Various types of dielectrics tend to form different amounts of fixed charges at the interface. Al2O3 contains negative fixed charges with a density in the order of 1012 cm‑2, while the number of fixed charges in HfO2 is one order of magnitude lower. SiO2 exhibits a similar density as the one for Al2O3, despite this they are of positive nature. One can make use of these particular features. By employing multi-oxide stacks, it is possible to influence the number of charges, which effectively act at the interface to Silicon. Fig. 1 shows the adjustability of the fixed charges between 3 x 1012 cm-2 and almost zero by adding a thin interfacial layer of HfO2. A SiO2 interface goes a step beyond. It offers the option to change the type of the fixed charges from a negative to a positive value continuously. Fig. 2 shows an example of an Al2O3/HfO2 nanolaminate.
(II) Conductive passivation: The next step towards advanced PERC cells is the introduction of a conductive surface passivation. There are two major advantages. On the one hand, opening the classical passivation layer in order to contact the cell itself involves process steps, which consume high amounts of energy. These process steps would be obsolete. On the other hand, losses attributed to the lateral diffusion of charge carriers towards the local contact openings and recombination of charge carriers at the openings itself will be reduced. Therefore, the efficiency of the cell will be increased. There are four major criteria to be fulfilled. The contacts have to exhibit a very high transparency for visible light, a good passivation behavior, a good conductivity and need to be compatible to the existing PERC process flow (see Fig. 3). Since 2016 NaMLab is investigating oxide layer stacks as potential candidates for this purpose. The main concept consists of a passivating tunnel oxide (Al2O3) with thicknesses below 1 nm and a transparent conducting oxide (doped TiOx). Therefore, the result is an enhancement of the highly passivating Al2O3 developed at NaMLab previously.
The applied atomic layer deposition technique provides highly accurate growth control in the sub-nm range. It opens the possibility to tailor material properties of Al2O3-based nanolaminates for novel functionality in future solar cell concepts.
Dr. Matthias Grube
TU Dresden (IHM) (Germany), FAP GmbH, Dresden (Germany), Meyer Burger AG, Hohenstein-Ernstthal (Germany)