These days, the most advanced robotic prosthetics take their commands from the brain. And pretty soon, they may be drawing their power from juices in the brain. Cerebrospinal fluid, that is. Electrical engineers at the Massachusetts Institute of Technology are developing a novel platinum-coated fuel cell that runs off the glucose found in bodily fluids. Their specific aim is to implant the fuel cells in liquid pockets of the brain and use them to run low powered components in a neural prosthetic. They described a prototype this week in the journal PLoS One.
Cerebrospinal fluid fills the central canal of the spinal cord, the ventricles of the brain, and the spaces between it and the skull. It's main function is to protect sensitive neural tissue from impact, but it also happens to be rich in glucose, a molecule which our bodies use for energy. The way we metabolize glucose is pretty complex, involving a cycle of oxidizing enzymatic reactions that knock electrons off the molecule one by one. At the end of the process, the molecule has been ravaged, all 24 electrons have been picked off and salvaged. Fuel cells work by the same principle, but only remove 2 electrons.
"Ours is like a baby oxidation," says Rahul Sarpeshkar, who led the research. "We don't have all the enzymes that the body has." Instead of enzymes, his lab uses a platinum coated anode to catalyze a reaction with the glucose in the cerebrospinal fluid, causing two electrodes and two hydrogen ions to pop off the molecule. The positively charged hydrogen passes through a membrane to a cathode on the opposing side, while the electrons get forced through an external circuit. Both recombine at the cathode tip, bonding with ambient oxygen to form water. At its most efficient, the circuit can generate up to 180 μW cm−2 of power, according to the study.
This isn't the first glucose fuel cell, and it's certainly not the only way to do it. Other groups, such as this one, coat their electrodes in enzymes that actively strip electrons from the glucose. This technique mimics natural metabolism, but it lacks a system for replenishing the metabolic proteins. "One of the problems with glucose fuel cells is that if you put in an enzyme that is efficient at oxidizing glucose, eventually that enzyme stops functioning," says Sarpeshkar. Once an enzyme runs out on a fuel cell, the whole thing has to be replaced. He expects a longer life span (though perhaps lower efficiency) for his own fuel cells. Because its only role is to catalyze oxidation, the platinum anode never runs out and never has to be replaced.
Another problem that a lot of implanted devices of all kinds run into is the immune response from the body. Over time, cells begin glomming on to the invader in a way that messes with the circuitry. But it turns out that the cerebrospinal fluid is a kind of forgotten zone in the body, nearly un-patrolled by immune cells. In fact, there are hardly any cells in it at all, making it a neutral little hiding place, perfect for docking foreign crafts.
Regardless of how the anode is designed, these fuel cells provide very little power and would be worthless without the ingenuity of those researchers who are adapting brain computer interfaces to consume less and less energy. As it so happens, Sarpeshkar leads the field on this front as well. According to him, his lab has claim to the world's lowest power nerve stimulator and neural amplifier. And in 2010 he published a book titled, Ultra Low Power Bioelectronics. When it's time to test Sarpeshkar's fuel cells in animals (for now they're working with artificial fluids), they will most likely be paired with one of his own medical devices.