PERSPECTIVE

 

This fuel cell technology contributes a system of integrating a power source on a logic wafer and removing many sub-functions and enabling more complex systems to be fitted into smaller spaces.  With the development of new, ever smaller, nano-scale electronic elements new solutions need to be identified with respect to routing power while isolating various types of unwanted signaling paths without involving costly additional processes.  This distributed power concept will provide a completely new opportunity for designing and realizing more versatile electronic devices.  The value of such a simplification is realized through the emergence of simpler implementation rules for future developments. 

 

System Architecture

One of the advantages of the semiconductor fabrication process is that the processing is extremely efficient in terms of cost per finished part.  This option does not presently include the electrodes of a fuel cell or other self-contained power source.  To produce a fuel cell in a standard semiconductor fabrication facility, it is necessary to produce a membrane using the current processes or compatible processes.  A membrane is required to be designed to reduce mixing of fuel and oxidant while permitting diffusion of positive ions.  This will be achieved by forming a region in which the membrane separating the fuel and oxidant is formed by the geometry and surface properties of the materials. 

 

Support of Logic

The fuel-cell potentials will bias their individual logic domains to the electrical operating region and will not require an electrical power regulation and distribution system.  The resulting chip, powered by distributed cells, can process data with extremely little electrical noise.  Additionally, it is possible to bring each domain power node to a pin and thus allow characterization or operation from an external power source.

 

The technology can be configured such as to provide:  

i.  Simple power distribution on complex chips,

ii. Extremely low noise,

iii.Little impact resulting from worst-case cycles.  

This can be supported by an HDL library, making the system commercially feasible by taking advantage of photographic mask integrity during automated fabrication and using local cells for local gating bias.  The architecture can be configured, so there is reduced need for components such as battery, regulator, chip isolation, and power distribution chain and thereby improving size, connection complexity and worst-case performance. 

 

The advantages of this innovation will result in the development of a fuel cell library with components that will enable straightforward integration of fuel cell elements with microelectronics.  Hardware Description Language (HDL) will be developed for the novel fuel cell elements, so the microelectronics designer can place them within a microelectronic architecture.  The complete logic and fuel cell combination can be produced in a standard microelectronic foundry.  The fuel cells will be deployed in a distributed manner within a microelectronic structure, such that the logic can flow between domains although the domains are powered individually.  This is a new fabric for design in which the power required per gate is much lower and the ground noise is reduced to near zero, leading to an increase in overall effective specific energy (i.e. operational time between refills) and/or enhanced capabilities of the microelectronic application.