Changing the sputtering system from planar to rotary targets increases the target utilization from about 30 to 75%, which saves expensive raw material. Depositing the molybdenum back contact by monolithic molybdenum rotary targets – which are completely made of molybdenum – gives additional benefits: The expensive bonding of the molybdenum material on a stainless steel backing tube is not required. In addition, a higher sputtering power of up to 30 kW/m can be applied. This creates a very high heat load which cannot be accommodated by bonded-type targets: Indium is used as bonding material which melts at only 156 °C thus the risk of de-bonding is eliminated with monolithic targets. A higher sputtering power results in a higher deposition rate and improved thin film properties, e.g. higher electrical conductivity.
In contrast to the high lateral homogeneity of thin films deposited by sputtering, a sputtered target has generally an erosion profile which is not homogeneous – a result of different plasma densities. As a consequence, the targets have to be replaced even if there is sufficient material on most of the target area. To overcome this limitation, targets with varying outer diameter (named “dogbone” for rotary targets) or thickness is a solution to extend the target lifetime and increase the target exchange interval.
Two elements have a big impact on the CIGS cell efficiency: iron (Fe) and sodium (Na). Whereas iron degrades the cell performance by introducing defects in the CIGS structure, sodium has the opposite effect: It concentrates defects at grain boundaries, which is a requirement for achieving high efficiencies.
Low iron levels are only achieved by a high purity of the primary materials as well as a high quality manufacturing process. Especially for molybdenum targets the iron content can vary significantly. The traditional way to introduce sodium into the CIGS absorber is by means of a soda-lime-glass substrate. During the manufacturing process sodium diffuses from the glass through the molybdenum back contact into the absorber layer, resulting in a sodium concentration in the order of 0.1 at%. However, reproducibility and lateral homogeneity of this process is low, and flexible substrates do not act as sodium source.
Solar cell producers now have an easy alternative at hand: By sputtering a layer of sodium doped molybdenum, the amount of sodium in the absorber layer can precisely be controlled and reproduced. PLANSEE offers MoNa sputtering targets with high purity and a uniform and fine grained microstructure. Tests in cooperation with the Swiss EMPA institute have already proven their benefits in practice: With sodium doped molybdenum layers the efficiency of CIGS solar cells could be significantly improved.