NASA NTRS2007
Elemental in-line evaporation on glass substrates has been a viable process for the large-area manufacture of CuInSe2-based photovoltaics, with module efficiencies as high as 12.7% [1]. However, lightweight, flexible CuInSe2-based modules are attractive in a number of applications, such as space power sources. In addition, flexible substrates have an inherent advantage in manufacturability in that they can be deposited in a roll-to-roll configuration allowing continuous, high yield, and ultimately lower cost production. As a result, high-temperature polymers have been used as substrates in depositing CuInSe2 films [2]. Recently, efficiency of 14.1% has been reported for a Cu(InGa)Se2-based solar cell on a polyimide substrate [3]. Both metal foil and polymer webs have been used as substrates for Cu(InGa)Se2-based photovoltaics in a roll-to-roll configuration with reasonable success [4,5]. Both of these substrates do not allow, readily, the incorporation of Na into the Cu(InGa)Se2 film which is necessary for high efficiency devices [3]. In addition, polymer substrates, can not be used at temperatures that are optimum for Cu(InGa)Se2 deposition. However, unlike metal foils, they are electrically insulating, simplifying monolithically-integrated module fabrication and are not a source of impurities diffusing into the growing film. The Institute of Energy Conversion (IEC) has modified its in-line evaporation system [6] from deposition onto glass substrates to roll-to-roll deposition onto polyimide (PI) film in order to investigate key issues in the deposition of large-area Cu(InGa)Se2 films on flexible polymer substrates. This transition presented unexpected challenges that had to be resolved. In this paper, two major problems, spitting from the Cu source and the cracking of Mo back contact film, will be discussed and the solution to each will be presented.