Are organic foods more nutritious than conventionally raised ones? Stanford University scientists cast doubt on that concept last year in a widely publicized report. But the gritty little secret is that whether your apples and spinach are organic or not, nutrient levels can vary dramatically depending on growing conditions, such as soil type and quality, temperature, and days of sun versus rain. As a consumer, you have no independent way of verifying that you have chosen a superior batch. But what if you had a handheld scanner that would allow you to check nutrient density? “You could compare carrots to carrots,” says Dan Kittredge, executive director of the Bionutrient Food Association, which is raising the funds to research such a device. “If this batch is a dud, pass. If the next one is good, that's where you spend your money.”

The basic technology has existed for decades. Near-infrared (NIR) spectroscopy—the modality that Kittredge is currently focusing on—has found applications in pharmaceutical manufacturing, medicine, agriculture and astronomy. NIR works on the principle that different molecules vibrate in slightly different ways. When infrared light is transmitted through or reflected from a given sample, certain wavelengths are absorbed more than others by the vibrating chemical bonds. By measuring the fraction of near-infrared light absorbed at each wavelength, scientists can obtain a distinct fingerprint that is characteristic of the sample. The results are precise—and fast. “Gas chromatography can easily take half a day,” says Magdi Mossoba, a research chemist at the fda's Center for Food Safety and Applied Nutrition. “NIR can give results in seconds.”

Until recently, NIR and related forms of vibrational spectroscopy were confined to the laboratory, where they required large benchtop instruments that only skilled scientists could operate. Now, with miniaturization, they are being packaged in simple handheld devices that “a worker without a Ph.D. in chemistry can use in a warehouse or in the field,” says Maggie Pax, a senior director at Thermo Fisher Scientific, a leading manufacturer of these tools. Pharmaceutical companies are using them to determine whether batches of raw ingredients are correctly labeled. More than a dozen countries have purchased them to help combat the rising tide of counterfeit drugs. And farmers use them to measure protein levels in grain, which helps to determine its market value.

Still, NIR has one major limitation as far as a supermarket scanner is concerned, which is that it cannot give readings for compounds at a concentration of less than 0.1 percent. The average vegetable is 92 percent water. After that come macronutrients such as carbohydrates and protein (in large enough quantities to be read by NIR), followed by micronutrients, including vitamins, minerals and antioxidants (most of which are too low to detect). The entire concept would be dead if not for one key observation. “Plants develop certain types of compounds in a predictable order and in specific ratios to various minerals, proteins and lipids,” Kittredge says. The task he is undertaking now with the Linus Pauling Institute at Oregon State University is to run thousands of assays on key foods to establish the algorithms needed to develop a workable scanner. “This will happen,” he asserts. “What we don't know is whether it will take three years or 30.”