7 March 2012—Drawing blood is a daily reality for most people with diabetes. And while checking glucose levels probably isn’t the worst part of the disease, it’s such a pervasive nuisance that someone from nearly every scientific discipline has tried to invent a better way to do it. They’ve pasted transdermal patches to the skin and shone near-infrared light through the earlobes, but still nothing can beat the accuracy of a little drop of blood.
Evidently, the quest for a better way is not over. Last month, engineers came up with new artillery—a plasmonic interferometer that can detect very low concentrations of glucose in water and, with some reengineering, may also work with saliva. If things go as hoped, people with diabetes will one day measure glucose levels by spitting instead of sticking.
Interferometers have been used in fields as diverse as astronomy and oceanography, and the devices come in all different sizes. In order to work as a glucose meter, they must be very small. “What we are doing here is bringing this concept down to the nanoscale,” says Domenico Pacifici, a professor of engineering at Brown University who led the research, which was published in a recent issue of the journal Nano Letters.
Each tiny interferometer consists of a slit in a silver film with two grooves etched on either side. When a beam of white light hits the film, the grooves in the silver scatter it in different directions. Some of the photons travel along the surface of the metal back toward the slit, and as they do so, they interact with free electrons in the material, swooping them up in a wave called a surface plasmon polariton. A wave originating from each groove surges toward the slit and collides with the light in the center, either increasing or decreasing the light’s intensity as it passes through the slit.
The measurement of this intensity gives an idea of how the light at the center interacts with the two interfering waves. “The slit is acting like a spatial mixer of those three contributions,” explains Pacifici.
In one configuration, with a bit of water dropped on the silver film, the interferometer will always give the same output. But if you change the distances between the grooves and the slit, or if you change the wavelength of the original light, the intensity changes on the other end. By trying many configurations, you can get a fingerprint specific to the water.
Because the interferometers are so small, the researchers can try multiple combinations at once, Pacifici says. “What we do is we etch thousands of those in the same chip in the same metal film. It allows us to have nanometer resolution,” he says. Then, when a molecule like glucose is added, it can be detected because it shifts the refractive index of the water, slightly altering its interferometric fingerprint.
This works for solutions like water, where the fingerprint has been worked out ahead of time. And saliva is 99 percent water. But the remaining 1 percent contains a mixture of unpredictable compounds, all of which will have their own influence on the refractive index of the solution.
Pacifici says he has not yet solved this problem, but he has some ideas of how to address it. One way would be to place a filter over the interferometer that would let only water and glucose through while leaving the rest of the saliva behind.
But the biggest problem is that the glucose level in saliva does not match up cleanly with the glucose level in blood, which is the measurement you really need to direct treatment. According to Guenther Boden, an endocrinologist at Temple University Hospital, in Philadelphia, it takes a long time for sudden changes in blood glucose levels to register in the saliva.
“It is not satisfying for a whole variety of reasons. The first one is that glucose equilibrates in these body fluids either incompletely or, if it does get in, there is certainly a tremendous time lag,” Boden says. “It’s very important that you know what’s the glucose right now—or maybe 5 minutes ago—but not much more than that.”
It’s an argument that Pacifici is familiar with. “Of course the time issue is a problem,” he says. But he remains unconvinced by research that has been done on the time correlations and says his device could at least be used to clarify how glucose levels in the blood and in saliva relate to each other.
Even if the technology does work, Boden is skeptical that there’s a serious need for it. He points out that glucose meters have improved to the point where they require only a tiny droplet of blood, which patients can get from patches on the body with a low density of nerves, like the underside of the arm. “It’s gotten so easy that it’s probably not worth spending millions of dollars to find a different way,” he says.
But Pacifici’s goals are not so narrow. His lab is also looking at ways to use the interferometer to detect signaling proteins called cytokines, which provide a measurement of inflammation in the body. “Surgeons are in need of a technology that can tell you in real time how much and how many of the cytokines are in the blood or in the serum so they can decide when it is best to operate,” he says.
So measuring glucose is only one proof for a device that could be used for many different applications, Pacifici says.
About the Author
Morgen E. Peck is a freelance writer based in New York City. In December she reported on the world’s first Bitcoin conference.