The advent of the artificial heart has spurred scientists to pursue synthetic kidneys and pancreases as well. Still, one key obstacle to realizing such devices is powering them after they have been implanted. Instead of having to constantly recharge them by hooking them up to some external system—or, worse, periodically removing them and replacing their batteries—researchers would prefer that these machines somehow harvest energy from their hosts.

Now there is hope that future implants might be powered not by batteries but by the fuels in our bodies that are used for energy. Scientists have shown that fuel cells implanted in rats can successfully generate electricity from sugar in the rodent's bodies. The devices kept going for months at a time.

The most potent sugar-powered fuel cells to date, so-called glucose biofuel cells, rely on enzymes that harvest electricity from chemical reactions—for instance, the combination of glucose with oxygen, both available in the human (and rat) body. Compounds dubbed "redox mediators" then act like wires, transporting electric charge from these enzymes to electrodes that lead from the fuel cell to whatever device it is powering. Scientists are currently pursuing a variety of such devices to generate electricity in an environmentally friendly manner.

Unfortunately, the enzymes used in past glucose biofuel cells were not suitable for implants, because they either required highly acidic conditions to work or were inhibited by a variety of ions found in the body. The newly developed devices lack these constraints and are the first functional implantable glucose biofuel cells, with prototypes in rats stably generating power for at least three months.

"It becomes possible to envisage development of implantable robots capable of compensating for failing functions in human beings," says researcher Philippe Cinquin, a biomedical engineer at Université Joseph Fourier in Grenoble, France.

Past glucose biofuel cells kept the enzymes and redox mediators in close proximity to electrodes by chemically bonding them on. However, not all enzymes and redox mediators lend themselves well to such bonding. Instead, Cinquin, with electrochemist Serge Cosnier and their colleagues, forego these bonds—they just physically pack enzymes and redox mediators into place on electrodes and then wrap the kind of membranes used in dialysis bags around them all. These semipermeable membranes allow fuel to flow in while keeping the enzymes and redox mediators from leaking out. That arrangement gave the scientists an opportunity to investigate enzymes that are more compatible with the body, an issue that had been neglected before.

Their most efficient prototype relied on composite graphite discs loaded with the enzymes glucose oxidase and polyphenol oxidase. Its two electrodes, which altogether took up just 0.266 milliliters of the 5-milliliter fuel cell, in total generated a peak power level of 6.5 microwatts, or millionths of a watt. (The current from the fuel cell fed into wires leading out of the body into hardware that controlled and measured the device, so it did not shock the rodents.) The scientists detailed their findings online May 4 in the journal PLoS ONE.

A milliliter's worth of these electrodes could in principle generate a peak power level of 24.4 microwatts, more than the 10 microwatts typically needed by pacemakers, the researchers added. The size of the devices matches well, too—pacemakers are typically between 10 milliliters and 25 milliliters in size, Cinquin says.

Given the success so far in rats and the fact that the prototypes are wrapped in plastics already clinically approved for implants, Cinquin sees no reason why they cannot work in people as well, and he hopes they will find a use in five to 10 years. A subsidiary company of his university, Floralis, will pursue industrial partners for this work, or perhaps launch a startup.

The first application of these fuel cells could be to power artificial urinary sphincters, which would require up to 200 microwatts, Cinquin explains. Currently 10,000 new patients each year suffer from incontinence after their prostates are removed, and their only recourse right now is a cumbersome pump inserted into their scrotums that patients have to press in order to urinate.

A more ambitious target would be an artificial kidney, which would require 20 milliwatts, or thousandths of a watt, to perform the same water-maintenance functions as kidneys do in people. An artificial heart is even further off, requiring at least several watts.

"The device they demonstrate brings a new level of performance to implantable biofuel cells," says chemical physicist and chemical engineer Plamen Atanassov at the University of New Mexico, who did not take part in this study.

However, "it is an open question as to whether their device can become more powerful to sufficiently revolutionize the field—I think they've pretty much taken it to the limit," Atanassov adds. The problem lies in the extracellular fluid such devices would get their fuel from in the body—the levels of the oxygen there are roughly 1,000 times less than the available glucose. "I don't think there's enough of an oxygen supply there," he says.

Cinquin is more confident they can improve the performance of the devices by tinkering with enzymes and redox mediators. In as yet unpublished work, "we have already improved performance by a factor of at least 50," he says. "Powering artificial hearts by glucose biofuel cells remains a long-term research objective."