To enable extraction of power from a solar cell, contacts need to be created to the negative and positive terminals of the device and conductive paths (usually made of metal) need to deliver the current and voltage out from the device. Hence, all solar cells have a metallisation process that fabricates such contacts and conductive pathways. Due to the large size of thin-film PV modules on glass, it is important to divide the large (~1 m2) initial thin-film solar cell into smaller unit cells and then interconnect them in series to keep ohmic losses at a tolerable level.

There are two reasons why pc-Si thin-film PV on glass can’t use the established method of monolithic series interconnection of the individual solar cells that is used in the a-Si:H thin-film PV industry for modules on glass/TCO substrates [78]: First, TCOs are not sufficiently temperature stable to withstand the high temperatures (>600 °C) used during some of the fabrication steps of pc-Si thin-film solar cells, eliminating the possibility to fabricate the pc-Si solar cell on a TCO layer (front TCO). Secondly, doped pc-Si layers have a much higher electrical conductance (i. e. a much lower sheet resistance) than doped a-Si:H layers and hence the thin-film cells would be severely shunted when a TCO film is deposited over their exposed side walls to connect the rear surface of one cell with the front surface of the neighbouring cell.

One method for metallising pc-Si thin-film solar cells on glass has been developed in recent years at UNSW [73, 79]. It involves two photolithography steps. The method is schematically shown in Figure 11.12. A thin (~100nm) SiO2 layer is deposited onto the entire rear surface of the solar cell, followed by the formation of a matrix of round openings (diameter ~30 |im, spacing ~80 |im, surface coverage ~5%) in the SiO2 layer. Various methods can be used to create these openings, for example via a conventional photolithography-based sequence involving wet – chemical etching (whereby no alignment is required for the photomask) or by controlled deposition of small droplets of hydrofluoric (HF) acid. The SiO2 layer is deposited via RF sputtering at room temperature. Next, an approximately 600-nm-thick Al layer is blanket deposited over the structure, for example using dc magnetron sputtering at room temperature. The purpose of this SiO2/Al stack is to provide both the rear electrode of the solar cell and a high-quality back-surface reflector (BSR). As outlined previously, the SiO2 layer boosts light trapping in the solar cell due to total






Figure 11.12 Schematic representation of the metallisation method developed at UNSW for pc-Si thin-film solar cells on glass. (a) initial structure; (b) before the plasma etching step for the emitter electrode (the openings in the rear contact stack were formed by photolithography and wet-chemical etching); (c) after the plasma etching step for the emitter electrode; (d) final structure
internal reflection at the c-Si/SiO2 interface. Next, a photoresist film is blanket deposited onto the Al film and patterned using a photomask with a conventional comb-like structure. This structure defines the location of the emitter (i. e. the glass-side) electrode. This patterning process requires no alignment of the photomask on the sample. A wet-chemical etching step then removes the aluminium and SiO2 below the openings in the photoresist film, thus locally exposing the pc-Si diode (see Figure 11.12b). Next, U-shaped grooves are etched into the Si film using a dry etching process (plasma etching) in a conventional 13.56-MHz parallel-plate plasma etcher, with SF6 as etching gas. The resulting structure is shown in Figure 11.12c. After a brief HF dip, a 600-nm-thick Al film is then deposited via e-beam evaporation. The photoresist and the overlying Al film are then lifted off by ultrasonic treatment in an acetone solution, giving the final structure shown in Figure 11.12d.

At UNSW, the interdigitated metallisation process described above is routinely used for the metallisation of four individual solar cells on 5 x 5 cm glass sheets. The metallisation process has proved to be robust and to routinely generate cells with fill factors of over 70%. The best FF realised as yet is 75.9% and has been obtained on a cell with an area of 4.4cm2. This is believed to be the highest fill factor ever obtained for a pc-Si thin-film solar cell on glass and is a clear proof of the potential of the method. The metallisation process has two photolithography steps, but neither requires alignment of the photomask. Since the UNSW pc-Si solar cells have a short high-temperature (>900 °C) defect anneal step which produces some glass deformation, precise alignment of photolithography masks would be cumbersome, slow and expensive. We also note that a low-cost LED array is used as UV light source for exposure of the photoresist, a method that can easily be scaled to large areas. Large-scale photoresist deposition via slit coating using micro­nozzles which spray the photoresist onto very large (>1m2) glass substrates is now a standard process in the LCD flat panel industry [80]. However, drawbacks of this metallisation method are that there is no PDR at the rear solar cell surface, that two separate aluminium depositions are needed, and that an extra processing step is required to series-connect the individual solar cells. Work is underway at UNSW to simplify the process, to incorporate a PDR, and to series-connect neighbouring solar cells.

Another method of forming a series-connected thin-film PV module based on polycrystalline silicon has been disclosed by Basore [75]. The technology is referred to as CSG (crystalline silicon on glass) and is the only pc-Si on glass technique that has entered industrial production. Device fabrication starts by using a pulsed laser to slice the Si layer into a series of adjacent, ~6-mm-wide strip cells. The module is then coated with a thin layer of novolac resin loaded with white pigments to make it more reflective and thus improve light trapping in the cell. Next the openings for the и-type emitter contacts (craters) are formed. This involves etching of openings into the resin layer (using an inkjet printhead), followed by chemical etching of the Si. Then the openings for the p-type rear contacts (dimples) are formed using the same inkjet process. A blanket deposition of sputtered aluminium provides electrical contact to the и+ and p+ Si layers. The aluminium film is then sliced into a large number of individual pads using laser pulses. Each metal pad series connects one line of p-type contacts in one cell with a line of и-type contacts in the next cell. The final structure is shown in Figure 11.13. It is noted that this metallisation and interconnection scheme again does not involve a TCO layer.

Advantages compared with the UNSW metallisation scheme discussed above are that only one metal deposition step is involved, that a PDR is incorporated, and that the solar cells are automatically series-connected. One challenge with the Basore technique [75] is the large number (millions/m2) of craters and dimples that need to be created. Another is that all craters and dimples need to be accurately positioned across the entire module, imposing significant challenges with respect to the alignment of the glass sheet and the patterning tools (such as inkjet or laser).

‘Crater’ ‘Groove’ ‘Dimple’

Textured glass

flight In

Figure 11.13 Schematic of CSG technology (from reference [75]; reproduced by permission of CSG Solar)