Wafers: Thermal Process Development

This is an excerpt from..

A Practical Guide for Improving Crystalline Solar Cell Efficiencies through
Optimization of the Firing Process

by Bjorn Dahle, KIC

As opposed to more mature industries such as semiconductor and electronics assembly, the solar industry currently does not have a clear understanding of the ideal wafer profile or process window for each unique application (wafer, silver paste, furnace and other variables). There may be numerous wafer properties and variables in the processes upstream from the metallization furnace that affect the ideal wafer profile. Therefore, each manufacturer must perform a design of experiment (DOE) to identify the best wafer profile or range of profiles (Process Window). Such a DOE typically involves changing the wafer profile while the cell efficiency, fill factor and other quality measurements are measured. The profile is changed numerous times (on identical wafers that have been processed up to the firing furnace) in a “trial and error” approach until the responsible engineer is satisfied with the cell efficiency.

The wafer profile itself is a result of the furnace settings and how the thermodynamic properties of the furnace heat and cool the wafer. A modern metallization (firing) furnace can be set up using tens of millions of alternative recipes (combination of zone temperatures and conveyor speed), making the DOE both difficult and time consuming. Even worse, when changing one profile parameter, e.g. the peak temperature, all the other parameters such as time above 500°C, time above 600°C, ramp rate, etc. also change. As a result, it becomes difficult to determine what caused the improvement in the cell efficiency.

Finally, the traditional methods to record the wafer profile suffer from inaccurate and non-repeatable measurements. This problem is caused by the method used to attach the thermocouples (TC) to the wafer. It is not uncommon to get a 50°C difference in peak temperature readings from one profile to the next when taken only minutes later. Clearly, a new TC attach method is required before any purposeful DOE can take place.

For more info go to www.kicthermal.com and www.beltfurnaces.com

Like Torrey Hills on facebook for updates on whitepapers, etc.

http://facebook.com/torreyhillstechnologies