Microcrystalline silicon and micromorph tandem solar cells H. Keppner 1 , J. Meier 2 , P. Torres 2 , D. Fischer 2 , A. Shah 2 1 University of Applied Science, 7 avenue de l’Hôtel-de-Ville, CH-2400 Le Locle, Switzerland (E-mail: keppner@eicn.ch) 2 Institute of Microtechnology, University of Neuchâtel, A.-L. Breguet 2, CH-2000 Neuchâtel, Switzerland Abstract. “Micromorph” tandem solar cells consisting of a microcrystalline silicon bottom cell and an amorphous sil- icon top cell are considered as one of the most promising new thin-film silicon solar-cell concepts. Their promise lies in the hope of simultaneously achieving high conversion ef- ficiencies at relatively low manufacturing costs. The concept was introduced by IMT Neuchâtel, based on the VHF-GD (very high frequency glow discharge) deposition method. The key element of the micromorph cell is the hydrogenated mi- crocrystalline silicon bottom cell that opens new perspectives for low-temperature thin-film crystalline silicon technology. According to our present physical understanding microcrys- talline silicon can be considered to be much more complex and very different from an ideal isotropic semiconductor. So far, stabilized efficiencies of about 12% (10.7% indepen- dently confirmed) could be obtained with micromorph solar cells. The scope of this paper is to emphasize two aspects: the first one is the complexity and the variety of microcrystalline silicon. The second aspect is to point out that the deposition parameter space is very large and mainly unexploited. Never- theless, the results obtained are very encouraging and confirm that the micromorph concept has the potential to come close to the required performance criteria concerning price and ef- ficiency. The justification of worldwide, intensified research on thin- film solar cells is not based on a lack of success in the efficiency performance of wafer-based silicon or GaAs tech- nologies. It is based on a lack of hope in the cost reduction potential of wafer-based technology. Indeed, in order to ren- der photovoltaics a competitive energy source in future, it is imperative to adopt the approach of low-cost solar cells. However, the abandoning of the “safe” wafer, with its phe- nomenally high diffusion lengths of hundreds of micrometers, has so far been, without any exceptions, always accompa- nied by a significant loss in efficiency. The efficiency drops down to less than a half, when compared to the record effi- ciencies of 24% achieved with wafers [1]. Nevertheless, the attraction of all thin-film concepts is based on the fact that the (generally expensive) semiconductor can here be deposited directly on low-cost large-area substrates. Furthermore, in all thin-film concepts, the cell must not be self-supporting and its thickness can therefore be chosen based on the absorption requirements. Thin-film solar cells based on compound semiconductors such as CdTe and Cu(In, Ga)Se 2 (CIGS), have attracted much attention in the past due to the remarkable work of many groups [2–4]; silicon-based thin-film solar cells were until recently exclusively limited to activities related to amorph- ous silicon (a-Si:H). Amorphous silicon technology has now achieved an industrial level [5, 6] and is economically com- petitive, contributing thereby to a reduction of the price per W P . However, a-Si:H has always been associated with low ef- ficiencies and with further efficiency losses during operation due to the Staebler–Wronski effect (SWE). Still, the low de- position temperature of around 200 ◦ C and the application of the monolithic series connection technique for module manu- facturing [6, 7] were generally considered as key features in order to obtain low manufacturing costs. Polysilicon material deposited at high temperatures by the CVD process is basically limited by the difficulty in finding low-cost substrates that have, on one hand, the same thermal expansion coefficient as silicon, and on the other hand, do not contaminate the growing layer [8]. Furthermore, CVD or PECVD processes for the deposition of thin crystalline sili- con layers will in general lead to layers with gap states and defects at the grain-boundary zones. These gap states act, as recombination centers, and, furthermore they screen the in- ternal electrical field necessary for carrier collection within p-i -n-type solar cells [9]. In particular, for solar cell appli- cations, the ratio of collection length to cell thickness (given by the absorption requirements) must be as large as possible. Note that due to the many grains and grain boundaries present in crystalline thin-films this ratio appears at first sight to be less advantageous for polysilicon than for amorphous silicon; hereby, the particularly strong absorption of a-Si:H overcom- pensates the poor electronic quality of this material. Hydrogenated microcrystalline silicon (μ c-Si:H) was originally introduced by Veprek et al. [10–12] and is nowa- days generally obtained by a PECVD process using a mixture of silane and hydrogen. Due to its efficient doping properties, both for n-type as well as p-type material, μ c-Si:H was from the beginning successfully used as ohmic contact layers in so- Published in Applied Physics A: Materials Science & Processing 69, issue 2, 169-177, 1999 which should be used for any reference to this work 1