Not falling through the cracks

Within the scope of solar module manufacturing, the wafer segment accounts for about 60 percent of the total costs incurred. Given such a significantly large portion of the cost-pie, it is pertinent for these wafers to be substantially error-free. Having said that, the efficient handling of these ultra thin elements in the production process of solar cells is an important challenge for solar cell manufacturing. As stated before, the ultra-thin characteristic of the wafers, apart from their value, makes them extremely fragile, with typical thicknesses being 0.18 to 0.20 millimeters (mm). This, thereby, translates into potential further cost penalties. Wafers have the tendency to break very easily should they be not treated in a gentle manner. The problem with that (apart from the fact that the manufacturer has broken wafers that become redundant) is that this may result in the decreased optimal utilisation of the production line significantly and furthermore lead on to the waste of costly production material. When examining current production costs and wafer prices, the increase of wafer breakage by one percent (e.g. from three to four percent) leads to costs about 500,000 euros per year for a 80 megawatt per year production line. These are the typical figures existent. That is already a hefty figure for a manufacturer to take into account.

Handling

Cell factories are equipped with handling and processing equipment, which is designed to handle wafers that remain within certain parameters with respect to geometry and surface properties, and most importantly rigidity. Any wafer unable to comply with these specifications has then the increased likelihood of breakage. Even if such a wafer with non-complied parameters survives the cell production process and becomes a solar cell, it has the potential to break in the module assembly segment, thereby causing even greater damages. Electrical properties of wafers can be checked with statistical methods. The parameters that are linked to breakage or shunts, on the other hand, need to be inspected 100 percent. At the same time, this needs to be undertaken without increasing the likelihood of breakage through the inspection process itself. Being non-tactile, fast and flexible, optical in-line wafer inspection is the established method for the incoming inspection of wafers in a solar cell factory.

Defects

There are a smorgasbord of defects that can come into play, and in order to match that, a wide array of tools within the equipment. These include geometry, box-warp, waviness, thickness, edge defect, saw mark and non-visible crack detection tools, to name the main ones. Wafer size and thickness (TTV = Total Thickness Variation) must stay within tolerance boundaries that are in the order of 0.1mm, usually not a big challenge for an optical in-line inspection system. Defects that occur at the outer contours, surface chips or the so-called saw-marks have to be detected as of 0.01mm. These are already a bit of a challenge. Saw-marks are usually caused by a sawing wire’s temporary deviation from its position. This thereby exhibits itself as steps in the wafer’s surface profile. The biggest challenge, however, is the detection of defects hidden within the bulk of the wafer. These are the so-called microcracks and inclusions of contaminations. Microcracks cannot be detected with visible light and may not even necessarily penetrate into the surface of the wafer. It is not entirely clear how these microcracks occur. Their occurrence seems to be related to thermal stress in the production process or physical stress during transportation and is directly related to the breakage rate in the cell production process. Given the wide variety of different inspection requirements, it is virtually impossible to combine all of them into one inspection system. In a typical system, the focus is normally on a sub-set of the inspection criteria. Some equipment even specialize specifically in microcrack inspection only.

Main requirements of systems

With the purchase of optical inspection systems, there is the expectation that these systems will make reliable statements with respect to the occurrence of defects, their size and type. Over- rejects, which means the system classifies acceptable wafers as rejects, should be minimal. Under-rejects, meaning the system lets non-acceptable wafers slip through, should not occur at all. In microcracked wafers, however, users prefer a low over-reject rate to avoiding under-reject by all means. Over-rejects cannot be sorted manually as the cracks are not visible to the naked eye. Trashing wafers, where there is uncertainty whether they are microcracked or not, is very painful. On the other hand, under-rejects would increase the breakage rate slightly but not influence the average cell quality significantly. Optical inspection systems are composed of a camera, illumination, and a system that presents the wafer to illumination and a handling system. The illumination is usually designed in a manner that the features that are to be inspected appear enhanced while those that are of no interest are at best not visible at all. This can be further elaborated as such. The wafer’s outer circumference, for one, is best inspected with the wafer being illuminated from the back. Surface defects are typically seen in a structured light (line projectors). Microcracks can appear as contaminations when the wafer is transparent and the light is penetrating the wafer in a direction parallel to the surface. Light sources for microcrack inspections operate therefore typically at a wavelength of 1.5 ?m (microns) and above. (That implies that cracks smaller than 1.5 ?m cannot be seen). In order to meet and deliver the reliability criteria as stated above, the wafers must be presented in the most predictable and controlled way to both light and camera. Transferring a measurement principle which works fine in the laboratory to the shop-floor is the most critical part of the system design. In order to control this step 100 percent, some system manufacturers supply a complete unit which includes the handling system. Others supply a tool kit which allows integration into the most popular process equipment. Solar cell manufacturers often contemplate retrofitting optical inspection systems into their existing production lines. Thus, with a special focus, the integration aspects have been covered in the survey.
It is also important to keep in mind that wafer manufacturers and cell manufacturers may have different requirements pertaining to the systems. Solar cell manufacturers are not satisfied with a system that aids in the removal of wafers with defects out of the production process only. Since a production process is optimized by carefully balancing the recipes of all production steps with respect to one other, it is crucial to export data from each process step into an off-line computer for analysis by the process technology engineers. The inspection system is used for “in-line characterisation”. Thus, inspection systems are expected to provide measurement data in a non-proprietary but standardised format (MS excel, MS Access, Oracle, SQL) using a standard data interface such as OPC or SEGS/GEM. Some systems even offer the option to define the classification criteria in database query formats. This thereby, allows the definition of the classification criteria following a careful off-line analysis of the measurement data.

Innovations and expectations

The industry has seen the introduction of some microcrack inspection systems into the market in the last two years. These systems are gaining a reputation for reliability and are about to be widely accepted. With the difficulty of microcrack inspection on wafers being overcome, the industry is anxiously waiting for reliable inspection systems with the ability to find microcracks in processed solar cells. Only such systems can potentially save the module manufacturers from the issues of breakage brought on by microcracks.

###MARGINALIE_BEGIN###

###MARGINALIE_END###