Big lab on a tiny chip.

■ Ultimately, consumers could have labs-on-chips in their home for fast diagnosis of common illnesses. —The Editors Imagine shrinking the beakers, eyedroppers, chemicals and heaters of a chemistry lab onto a little microchip that could dangle from a key chain. A growing number of companies and universities are claiming to have devised such marvels, ready to perform vital analyses from detecting biological warfare agents in a soldier’s bloodstream to identifying toxins in a tainted package of hamburger meat. Almost all the new devices are surprisingly far from portable, however. The sensor that examines a drop of blood or speck of beef might indeed fit in one’s hand, but the equipment required to actually move a fluidized sample through the chip’s tiny tubes often occupies a desktop or more. Two research teams are overcoming that hurdle with creative microfluidics—the precise manipulation of microscopic droplets. By moving liquid molecules with air or electricity, the groups are integrating the equipment needed to sample, analyze and report, all on a fob the size of a USB flash drive. And although the current chips are being crafted by hand, the designs could ultimately be mass-produced. That prospect would finally bring labs-on-chips to the places they are most desirable—the developing world, the battlefield and the home—where they could quickly detect HIV, anthrax or Escherichia coli. A chip could even be implanted into a diabetic’s body to help monitor the person’s glucose and insulin levels. Pushed by Air As a tool, labs-on-chips have become increasingly popular among researchers because they can conduct hundreds of experiments simultaneously at a mere fraction of the time, space and cost of longstanding benchtop processors. Tiny channels and valves inside the chips can heat, cool or mix small samples and reagents, as well as enable more exotic tests such as electrical stimulation. The surrounding apparatus required to perform the internal heating, cooling and mixing is often comparatively bulky, however, because of the unique behavior of fluids. When trapped inside incredibly narrow tubes, even watery compounds behave like syrup: they are difficult to push around. And ironically, when they do flow, they are remarkably free of turbulence, making it hard to mix them with reagents for chemical reactions. Forcing liquids through the chips with compressed air requires bulky plumbing. Electrically driving the liquids requires high-voltage power supplies. The tests these desktop labs-on-chips can handle have gotten progressively more sophisticated. In 1998 chemical engineer Mark Burns and geneticist David Burke of the University of Michigan at Ann Arbor demonstrated the first chip that could identify a particular gene or variation of it. Since then, the researchers have INNOVATIONS