During diffusion, the entire surface of the wafer is exposed to the dopant source, including the rear of the solar cell and edges. In the case of a phosphorous diffusion, this creates a current path from the front junction to the rear of the device, effectively shunting the solar cell as these recombining carriers do not contribute to the power output. Therefore, following diffusion, an edge isolation process is required to remove the unwanted diffusion around the edges of the solar cell, and electrically isolate the front and rear surfaces. This can be done in one of three ways:
In this process, the samples are stacked on top of one another such that only their edges are exposed. They are then loaded into a chamber where they are exposed to a plasma formed of CF4 and O2 gasses, which etches the exposed sides of the wafer, effectively removing the diffused silicon at the edges. A short process results in insufficient isolation and therefore high shunt resistance. In contrast, an extended process may result in excessive damage to the edges which damage the diffused junction, which enhances recombination in this region, affecting the solar cell ideality.
Wet chemical etching
In this process, the samples are etched using a mixture of hydrofluoric acid and nitric acid on one side only, in a single-side etch tool. This method ensures the complete removal of the rear diffusion and edges, without the risk of damage ion damage from a plasma etching process. For this reason, in addition to the in-line nature of single side etch tools which enables high throughput and low breakage risk, the wet etching process is the most common method used in production for edge isolation. However, one risk with this process is the potential for the etchant to splash onto the front-side, and so the process needs to be controlled and monitored carefully.
Laser edge isolation
In this process, a laser is used to ablate a trench through the front surface of the wafer close to the edge, effectively breaking the current path to the edge. Similarly, dopants of the opposite polarity may be applied to the front surface (e.g. via a spin-on dopant source) and diffused into the wafer via locally melting the sample with a laser, which effectively incorporates the dopant into the silicon upon recrystallisation during cooling. This method creates a diode around the edge of the wafer, preventing lateral carrier flow to the edges. Unlike the other two methods, laser doping edge isolation is performed after co-firing the silver contacts.