海角大神

It鈥檚 a staple of science fiction, made famous by the tricorders on 鈥�Star Trek鈥�: a hand-held device that reveals detailed information about some unknown substance or object in front of you. Sometimes you鈥檇 even get a real-time picture of each molecule.

Laboratory devices can decipher such unidentified things, but the special equipment is often bigger than a refrigerator and many times more expensive. Not exactly tricorder technology.
But David Grier at New York University believes that he鈥檚 closer than ever before 鈥� and his design uses parts available from a local electronics store.

The setup is simple: a laser, a microscope, a digital video camera, and a PC. Take the laser and fire it through the microscope 鈥渂ackward鈥� 鈥� from behind the object you are looking through the lens. The image that hits the microscope looks like a pattern of rings, like ripples in a pond. With a little computing power, Dr. Grier can read the pattern of circles and create a real-time image that teases out the defining characteristics of an object.

With an ordinary microscope, you can only see a two-dimensional image. But the ring pattern made by the laser allows the user to measure how far the object is from the lens. Since different materials refract light in different ways, you can tell exactly what the target is made of.

The analysis works on liquids, goos, and dusts 鈥� things translucent enough to allow laser light to pass through 鈥� but also solid objects. This 鈥渞ipple鈥� effect is just barely visible in ordinary light. It鈥檚 what creates the 鈥渇uzzing鈥� effect at the edge of shadows. It鈥檚 also visible at sunset 鈥� brilliant red sunsets are due to dust scattering longer wavelengths, and one could use that to determine the average size of the dust particles. Grier鈥檚 technique allows for observing smaller samples in a more controlled way and is accurate with remarkably small samples: down to sizes measured in micrometers, or millionths of a meter. A coat of paint is typically 100 micrometers thick.

鈥淭he one place we did go more all-out and used professional equipment was the microscope,鈥� Grier says. 鈥淭he lenses are higher quality.鈥� The rest of the equipment, he said, isn鈥檛 any different from what鈥檚 available off-the-shelf.

In the lab at NYU, Grier鈥檚 partner in research, Fook Chiong Cheong, shows the setup, which takes up three feet of a lab table.

鈥淚n the next generation of the device, we can make it smaller,鈥� Mr. Cheong says. He pointed to a spare video camera lying next to the apparatus. 鈥淭hat one is color,鈥� he says. 鈥淭he old one is monochrome.鈥�

A color camera enables the user to see exactly how varying wavelengths scatter, giving even more information about whatever it is on the slide. The new camera, he adds, is also cheaper.

Unlike other methods, this one doesn鈥檛 damage the object in question. On top of that, there is no longer any need to 鈥渢ag鈥� substances with radioactive markers or fluorescent compounds, which is expensive and time consuming.

Grier said the next generation of his invention should be able to fit into a soda can. And by adding some fiber-optic cables 鈥� also available at the local Radio Shack 鈥� he can use lasers of multiple wavelengths, increasing the sensitivity of the device.

Why didn鈥檛 anyone think of this before? The basis for imaging this way is called Lorenz-Mie theory. It鈥檚 all based on a single equation, which has been well known for a century. But doing the calculations to reconstruct the position, velocity, size, and refractive index of the object is complicated at best. Grier notes that on older machines, the calculations can take a long time.

鈥淏efore, you鈥檇 have been old before the calculation was done,鈥� he jokes. That all changed with the advent of more powerful PCs and digital imaging technology.

鈥淧eople had been satisfied for years with 鈥榖ack of the envelope鈥� calculations,鈥� he says. Those older attempts mostly produced snapshots, not precise images.

Developing the new software was the tough part. Grier says the problem is that there were too many moving parts. The laser produces a scatter pattern that changes with time, and you have to track the target鈥檚 position, refractive index, and size from the constantly changing image. That鈥檚 a total of six parameters (position is three for three dimensions, plus velocity).

At first, Grier鈥檚 team tried to standardize two or three of those numbers in the hopes that the others would then come more easily. That didn鈥檛 work. 鈥淢ore or less out of resignation we just let all of them vary,鈥� he says. 鈥淭hat worked.鈥�

Joseph Katz, professor of mechanical engineering at Johns Hopkins University in Baltimore, says the big difference in Grier鈥檚 technique is the ability to study dynamic systems, instead of having to deal with the narrow depth of field an ordinary microscope offers. (It was one of Katz鈥檚 original papers that inspired Grier鈥檚 work.)

Because you can tell where small particles are moving, you can track how they diffuse through other substances. One recent experiment 鈥� designed by a student 鈥� involved eggs.

鈥淭he administrator looked at us a little funny when we said we needed to buy eggs,鈥� Cheong says.
The team used organic eggs, commercial eggs, and even ostrich eggs (available at the local Whole Foods supermarket). The result? Certain kinds of particles diffuse less well through organic eggs. They haven鈥檛 figured out why that is yet.

Another application is dentistry. Cavities happen because bacteria get on your teeth and live in a film that sticks on the surface. If the film could be analyzed and disrupted, the bacteria would die off.

Cheong notes that because the imager can 鈥渟ee鈥� very small particles, it can show exactly how nutrients affect the bacteria.

鈥淚t turns out the bacteria feed on both sugar and starch,鈥� he says. 鈥淪o it鈥檚 worse to eat cake than it is to eat candy.鈥� The dream? Use results from this device to create a chewing gum that would eliminate the need for toothbrushes.

One of the reasons Grier started the work was frustration with conventional imaging. Much of the work in his lab deals with finding out how very small particles interact, as well as how to manipulate them. He wanted, he says, to 鈥渟ee what he was doing.鈥�

A similar laser setup can be used to make 鈥渓ight traps鈥� that use focused laser light (slightly different from a laser beam) to hold individual particles in place. The technique is well known, but Grier鈥檚 imaging technology lets experimenters see the results as they happen.

With a setup not unlike a video game, Cheong shows how to move tiny glass beads into place, using just a mouse and looking at the image on a screen. So far, there haven鈥檛 been any 鈥渟erious鈥� applications. But for Cheong, that isn鈥檛 a problem.

鈥淲e try to encourage play here,鈥� he says, adding that it is where good ideas sometimes come from.