Silicon Based Direct Methanol Fuel Cells

The purpose of this project has been to investigate and fabricate small scale Micro Direct Methanol Fuel Cells (μDMFC). They are investigated as a possible alternative for Zinc-air batteries in small size consumer devices such as hearing aids. In such devices the conventional rechargeable batteries such as lithium-ion batteries have insufficiently low energy density.

Methanol is a promising fuel for such devices due to the high energy density and ease of refueling compared to charging batteries, making μDMFC a suitable replacement energy source. In this Ph.D. dissertation, silicon micro fabrication techniques where utilized to build μDMFCs with the purpose of engineering the structures, both on the micro and nano scales in order to realize a high level of control over the membrane and catalyst components.

The work presents four different monolithic fuel cell designs. The primary design is based on a perforated silicon plate which acts as a mechanical support structure a proton conducting polymer membrane, which connects catalyst layers deposited through spray coating on either side of the silicon device.

An improvement of this design is also presented which integrates the catalyst layer into the current collector electrodes. This design is based on catalytic in situ growth of carbon nanotubes and atomic layer deposition of active catalyst particles. The additional two fuel cell designs utilize a porous silicon structure as the mechanical support, using respectively a spray coated catalyst and atomic layer deposition for.

This method of integration was investigated as a high internal volume support structure with potential for rapid batch fabrication. In characterization of the devices the work presents the development of an electrochemical impedance spectroscopy measurement setup capable of determining the ionic conductance of the devices.

Along side this a simple model for the ionic conductance is developed. A theoretical framework for the device performance is developed, and amongst the chief theoretical results are the determination of the tradeoff between conductivity and cross-over which results from variation in the path length through the proton conductive phase.

In addition the trade-off between mass transfer losses and activation losses deriving from the catalyst layer density is developed.