Search

27th European Photovoltaic Solar Energy Conference and Exhibition; 1064-1067, Economical applications of photovoltaic require constant efforts in the reduction of the crystalline silicon wafer cost. Approaches for a cost reduction of PV devices are inexpensive raw material sources (e.g. um-Si, remelt,…), or a better utilization of the raw material for example by cutting thinner wafers (i.e., more wafers per crystal length) and reducing the loss of raw material during slicing (saw dust or “kerf” loss). Nowadays, the race toward high-efficiency (+20%), low cost commercial solar cells based on multi- or single-crystal Si technology is mainly focused on thin and large wafer manufacturing capability. This is because the wafer thickness is an essential cost driver and a pristine parameter for controlling the efficiency in high-efficiency cell designs. With cell thickness today in the range 180μm- 220μm, and global PV roadmap aiming at even thinner wafers for accrued cost savings, we investigated the capability of the multi-wire slurry slicing process to further reduce the wafer thickness. In particular, we report on the impact of decreasing the wafer thickness onto the wafer geometrical characteristics such as Total Thickness Variation (TTV). A simple first-order model for TTV is presented yielding insights on potential routes to improve the overall wafering yield.

27th European Photovoltaic Solar Energy Conference and Exhibition; 1010-1013, In this paper we studied the TTV (Total Thickness Variation) dependence on different metrology setups. Typical wafers cut by multi-wire saw show maximum thickness on one edge and minimum thickness on the opposite one. TTV is mainly due to the thickness of the 4 wafer edges, hence sensor locations and edge exclusions are decisive in TTV measurements. In this article we present a comparison of different TTV values measured by sensors located in different positions relative to the lateral edges of the wafer with the choice of top and bottom edges exclusion matching the lateral distance. The measurements have been carried on utilizing the same lot of wafers. The result is highlighting the need to have sensors positioned exactly at the same distance with respect to the lateral wafers edges and also the need to set the same top and bottom edge exclusion in order to compare TTV measurements from different systems.

27th European Photovoltaic Solar Energy Conference and Exhibition; 957-959, The surface quality of silicon wafers cut by multi-wire saw must be controlled in order to reduce the breakage risk during the process as well as to improve the final solar cell efficiency. In addition, investigating the wafer surface can also provide useful information for improving the wire-sawing process. In this paper, the experience with running wafer inspection lines and a wire sawing test lab will be described. The objective is to demonstrate the correlation of the wafer surface defects with the wire saw parameters, as well as to stimulate discussion on the use of wafer inspection systems for wire-saw process improvements. A database with more than 200 different cutting conditions was investigated. Certain types of surface defects in this investigation were found to be strongly related to the cut parameters variations during the cut. Building up a centralized system to monitor several wire saw machine parameters and the corresponding wafer inspection data contributed to manage the operating status of the machines as well as to improve the cutting process, which can eventually lead to cutting yield improvements and wafering cost reductions.

27th European Photovoltaic Solar Energy Conference and Exhibition; 960-963, Wire sawing is the most common technology to fabricate silicon wafers in the current solar cell industry. The sawing performance is determined by operating parameters such as the wire speed, the table speed and the wire tension besides the wire types and the wearing status. In order to enhance the cutting performance as well as to maintain the mechanical stability, those parameters are usually adjusted by a predefined recipe along the cutting process. In this study, the material removal process was numerically simulated along the cutting time. The simulation was used to identify the low cutting efficiency regime. The adjusted recipe has been proved by experiments to demonstrate the validity of the numerical modeling.

26th European Photovoltaic Solar Energy Conference and Exhibition; 1955-1958, A new numerical model was developed to compute the total cost of ownership (CoO) considering the physical behaviors of control parameters associated with wire sawing. A parametric analysis was conducted to find an optimal operating condition to reduce CoO by implementing the developed model in a mathematical program, MATLAB. The variation of each contributor to CoO such as silicon, wire and slurry costs was identified in detail. The sensitivities of CoO versus a few primary parameters were compared and showed that the two major contributors on CoO are the poly-Si price and the slicing yield.