Professor Rahul Sarpeshkar and colleagues at MIT have created an implantable fuel cell which relies on glucose as its fuel. The device could potentially be used as a power source for the computers needed to decode brain signals and manipulate prosthetic or perhaps paralyzed limbs. The article was published in PLos ONE.
There are many layers of cool in this story. Implantable glucose fuel cells have been invented before – back in the 70s – but contained enzyme-based anodes which degraded over time and needed to be replaced. Because of this, implants like pacemakers rely on lithium ion batteries – which also drain over time, but have a much longer lifespan. This design utilizes a platinum anode to oxidize glucose ultimately to gluconic acid, liberating 2 electrons. The cathode is a matrix of single-walled carbon nanotubes which reduce dissolved oxygen to water.
What strikes me as the coolest part of this story, the fuel cell is fabricated on a silicon chip and would be placed in the cerebralspinal fluid next to the brain. Platinum is already known for being fairly biocompatable, but placing the chip in the cerebrospinal fluid is beneficial as there are very few white blood cells in the CSF to trigger an immune response and potential rejection. The CSF contains roughly the same concentration of glucose as plasma, and is not predicted to consume enough glucose fuel so as to impair brain function.
To avoid short circuiting the fuel cell, the platinum anode is enclosed in a ring of the carbon nanotube cathode. The cathode sequesters the dissolved oxygen for the reduction reaction, creating a concentration gradient across the cathode such that dissolved oxygen does not penetrate past the cathode. The nanotubes do not oxidize glucose, but are permeable to glucose, so the glucose passes through the cathode unreacted. The cathode is separated from the anode by a strip of nafion, a biocompatable perfluorniated polymer similar to teflon, but with branched sulfonate groups throughout. The sulfonate groups allow protons to flow through the nafion from the anode to the cathode, and nafion is permeable to glucose as well. The anode is at the center and can be as large as 2.5cm x 2.5cm.
The oxidation reaction at the anode is considerably less efficient than other implantable glucose fuel cells, but the biocompatability and long lifespan of the fuel cell makes this a really nice step forward in the treatment of paralysis and spinal cord injuries.