Glucose energy harvester for self-powering of remote distributed bioanalytical microsystems

Abstract

Implantable medical devices came of age with the artificial pacemaker more than 60 years ago. Their true potential of being located directly where their prosthetic or therapeutic actions are needed will be further enhanced by making them smaller with the aid of microfabrication. Their Achilles heel comes at the cost of limited lifetimes of an otherwise functional unit due to their small footprint limiting the size and energy storage capacity of current power supply solutions based on batteries. In contrast, chemical energy harvesters based in exogeneous fuels such as glucose and dissolved oxygen may hold the best promise of developing a long-term energy supply due to the relative abundance of these fuels in vivo. This thesis focuses on the development of such a chemical energy harvester. Based on the depletion design published in prior art, the quest has been to design a glucose fuel cell with electrodes and catalysts that are compatible with thin film microfabrication. These are mounted as a stacked assembly where one of the electrodes is made as a novel porous cathode through which the reagents diffuse and the initial separation of reagents takes place. The first challenge of making an electrode for the selective reduction of oxygen was realised by developing a porous cathode from e-beam deposition of palladium thin films on ceramic aluminium oxide substrates. The porous nature of the cathodes improved the catalytic properties by increasing the real surface area close to 100 times of the geometric surface area. It yields an exchange current density of 2.9 × 10−3 ± 0.5 × 10−3 μA cm−2 at a dissolved oxygen concentration close to the physiological range of 2 ppm. The sensitivity towards glucose was assessed by measuring the decrease in the half-cell potential in the presence of 5 mM glucose to - 20.6 ± 16.1 mV under a load current density of 2 μA cm-2. The Tafel slopes were measured to approximately 60 mV per decade. These results suggested that nanoporous AAO cathodes coated with palladium offered a reasonable catalytic performance with a good selectivity towards oxygen in the presence of glucose. The second challenge of realising selective oxidation of glucose started off by developing glucose selective anodes from the annealing of e-beam deposited thin films of platinum (Pt) and nickel (Ni) into a Pt– Ni alloy. The roughened surface of the alloy enhanced the electrochemical properties by increasing the real surface area to approximately 500 times compared to the geometric surface area. Since the surface roughness was found to scale with the annealing temperature, the corresponding exchange current density of the electrodes annealed at 800oC was twice that of the electrode annealed at 650oC. The potential increase due to the addition of dissolved oxygen at the physiological concentration of 2 ppm was measured to 100 ± 8 mV under a load current density of 2 μA cm-2. These results showed that the anodes are relatively more sensitive to oxygen catalysis than the cathodes towards glucose catalysis. It suggested that a shield should be made to remove the oxygen before permitting the solution to come into contact with the anode. Consequently, the third challenge of completing a selective oxidation of glucose and a selective reduction of oxygen lies in the architecture of the fuel cell device. A custom housing enabled a stacked assembly of the fuel cell “core” meaning the anode at the bottom, covered by an ion conducting membrane, and capped with the cathode, showed that energy harvesting from a mixed fuel environment was possible. In fact, the cell was able of maintain a power density of 2.33 ± 0.11 μW cm-2 at a current density of 7.7 μA cm-2 and a cell potential of 0.30 ± 0.01 V in a simulated mixed fuel environment of 5 mM glucose and 2 ppm dissolved oxygen at room temperature. This was 80% of the power obtained in the ideal experiment in which glucose and oxygen were physically separated prior to use. It was also found that the methods used to estimate the real surface area of nanoporous electrodes were not thoroughly discussed in literature. Although this may be clear to those already skilled in the art, mistakes can be done by those coming from a different engineering background. Thus a subsequent study investigating different popular ex-situ and in-situ methods was undertaken to help clarify this matter and to identify the correct methods used for the electrode systems developed in this project.