A research group at Massachusetts Institute of Technology (MIT; Cambridge) is getting a sweet taste of success upon developing a proof of concept for a biocompatible fuel cell capable of harvesting energy from glucose in the body. By achieving a delicate synergy between ultra-low-power electronics and highly energy-dense systems, the glucose fuel cell could someday eliminate the need for batteries and provide an unlimited, lifetime power supply to brain-machine interfaces, pacemakers, and a variety of other implantable medical devices.

"It's always been a Holy Grail in the medical device field--and low-power electronics in general--that if you create a device that is low power enough, you should just be able to run it off of harvested energy," says Rahul Sarpeshkar, an associate professor of electrical engineering at MIT and lead researcher on this project. "People are trying to build devices powered by vibration or body heat, but the natural energy harvesting you can do in the body is with glucose; it is a very energy-dense molecule and allows you to have a lifetime power supply for an implantable device."

Previous attempts at creating glucose fuel cells and batteries, Sarpeshkar notes, have been stymied by such issues as enzyme fouling, impracticality for long-term implantation, and an inability to match the power of traditional batteries, for example. Leveraging his lab's experience with ultra-low-power devices and bioelectronics, however, Sarpeshkar and his team decided to explore the possibility of developing a glucose fuel cell that exploits the use of platinum as a catalyst to power an energy-efficient brain-machine interface. The use of platinum, Sarpeshkar adds, serves to catalyze glucose oxidation, mimicking the role of enzymes in the body, albeit relatively inefficiently.

"We started this research as sort of a high-risk, high-reward project," Sarpeshkar comments. "We settled on implanting the glucose fuel cell in cerebrospinal fluid; it has good glucose content--although not as much as in blood--and also has very few cells and proteins that can cause biofouling." He adds that his former student, Benjamin Rapoport, found that the rate of glucose and oxygen consumption required to support an ultra-low-power implant in that area was sufficient as well.

To create the glucose fuel cell from silicon, the MIT researchers, in collaboration with researcher Jakub Kedzierski from MIT's Lincoln Laboratory, relied on basic semiconductor fabrication techniques. The team began with a platinum-aluminum alloy that it then roughened using a nanofabrication etching process to remove the aluminum. This roughening step, according to Sarpeshkar, serves to increase the surface area, which, in turn, improves catalytic efficiency of the platinum.

"More catalytic sites means more power," he explains. "The improvement in power then allowed us to meet our very-low-power specifications. If we had not done the roughening of the platinum, we would have been an order of magnitude below what the electronics could handle."

Combining ultra-low-power electronics with highly energy dense systems has ultimately yielded a fuel cell that could enable the development of a plethora of glucose-powered implantable medical devices. And while the prospect of commercialization is likely still several years away, the MIT researchers will continue to make progress, aiming to advance to animal studies in the near future.

"With a glucose fuel cell, you never need to change the battery in cardiac pacemakers or paralysis implants; there's a lifetime's supply of power," Sarpeshkar says. "There are also no concerns about battery safety, heating, or shorting, and you can save space because the fact that you're using glucose energy means that you have gotten there with a small, low-power device. Everything just gets better."