Abstract

The formation of highly doped p-type (boron-doped) and n-type (phosphorus-doped) poly-Si thin films with spin-on dopants on the same wafer poses unique challenges. This work evaluates the pros and cons of different approaches to dopant diffusion, residue removal, and diffusion mask selection when working with spin-on dopants. The experiments have been carried out to optimize the formation of highly doped poly-Si thin films that will form the active layers of a micro-thermoelectric generator (μ-TEG). Building upon previous work on residue removal methods, we report that the residue formation is also dependent on the oxide-silicon surface with a PECVD oxide-Si interface not requiring a post-dopant diffusion nitric acid treatment to achieve a hydrophobic surface. Furthermore, comparing the direct dispense and proximity diffusion methods verifies the existing reports of enhanced diffusion in the latter case. Besides, we observe that doping first with Boron is more conducive than doping first with phosphorus to reduce sheet resistance. We also optimize the contact formation with Ti/Ni thin films and report the mean contact resistivity. Notably, contact annealing precludes the nitric acid treatment requirement even if the surface is hydrophilic after removing dopant residue. Additionally, the sequence of silicidation is observed to be dependent on the doping of the underlying areas.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call