SAN DIEGO, Calif. A method for fabricating porous-silicon nanoparticles that have a selective response to light could lead to a fundamentally new capability for chemical sensing. Developed here at the University of California, the process creates a special reflective layer called a rugate filter on the surface of the nanoparticles.
The layer only reflects light in a narrow spectral band. The reflection bands vary, creating a medium that has a 20-bit code. A given chemical will change the reflectivity of a cloud of particles, creating a unique signature that can be detected from a distance.
"So far we have been able to detect the color changes in particles from about 20 meters away, but we estimate that we can use the same approach to detect them up to 1 kilometer away," said Michael Sailor, the inventor of silicon "gunpowder" at the University of California. Sangeeta Bhatia, an associate professor of bioengineer ing, also worked on the UC "smart-dust" project.
Smart dust began as a battlefield technology designed to facilitate testing for air- or waterborne toxins. The technology was used to fabricate microscopic-size particles and make them respond to specific chemicals. Several approaches have been tried already, such as using molecules with specific vibrational signatures, quantum dots or fluorescent particles.
Common among those diverse approaches has been functionalizing the surface of the particles so that they change in a detectable manner when encountering a qualifying toxin. So far, functionalizing the surfaces has been the easy part several methods now exist for creating "lock and key" mechanisms that flag specific toxins.
Sailor and his team functionalize the surface of porous-silicon particles in two ways. The first is a method of creating "size traps" for molecules of a specific size, which is useful for detecting a range of related particles like hydrocarbons. The second approach is to create traps specifically shaped for detecting molecules of individual chemical agents, or protein-encoded traps that bond only to specific toxic biological agents.
While functionalizing the particles for the detection of specific pathogens has been the easy part, the drawback with all approaches to smart dust so far, according to Sailor, has been sensing the changes in the smart dust from a safe distance. Measuring the changes in fluorescence, for instance, requires close-up monitoring of the particles, which partially defeats the purpose of detecting pathogens before they can affect personnel.
"Other approaches to smart dust are going to have a harder time sensing changes from a distance, but our particles are so highly reflective that we can sense changes from a distance just by using a more powerful laser," Sailor said.
The rugate-encoded particles could test for millions of possible pathogens from chemical to biological by functionalizing particles with different "colors" for different pathogens, then dispersing all the different particles together. Particles can be dispersed in air or water and then safely monitored from a distance by illuminating them with a laser and cataloging the reflected colors. Particles that have not encountered their chosen pathogens will retain their original color, but particles that have bonded to specific pathogens will have their colors shifted in identifiable ways.
Rugate filters were invented to create a narrow optical peak in the reflectivity of a mirror. Based on an "interference" coating, rugate filters are created by continuously varying the refractive index of a film in the direction perpendicular to the film plane. By varying the refractive index periodically between two extreme values, a very narrow band can be created in the reflectivity of a mirror.
Sailor's group encodes an entire wafer with the same reflective-color selectivity with a galvanostatic anodic etch of the crystalline silicon. The process creates an optically uniform layer of porous silicon, the thickness and porosity of which are controlled by the electrochemical process. By varying the current density during the etch, the duration of the etch and the composition of the etchant solution, Sailor's group was able to demonstrate that millions of specific colors (20-bit codes) could be selected for.
Identical color code
Current density was modulated during the etch with a sinusoidal signal. After the team etched the multilayered structure, they removed particles from the substrate by applying a current pulse simultaneously with agitation and ultrasound vibrations. As a result, thousands of micron-size particles of various sizes and shapes were broken off, but all had the identical rugate-filtered color encoding. They also successfully created more uniformly shaped particles by first patterning the chip's surface with lithography before etching.
"The multilayered photonic crystals we created display a very sharp line in the optical-reflectivity spectrum, much narrower than the fluorescent spectrum of molecules or quantum dots. Our full width at half maximum [FWHM] was only 11 nanometers, whereas reported quantum dots have an FWHM of 20 nm," Sailor said.
Sailor's group reports successfully preparing different particles with narrowband reflectivity in both the visible and near-infrared bands. In addition, the group has demonstrated that it can encode a single particle with more than one reflective band. For instance, the group prepared one set of particles that has three narrow bands of reflectivity, creating in effect a 3-bit code, where each "bit" can be coded for any of a million different values (colors).
Initial testing to try and detect the codes on those particles once they are released has shown that the method is only 90 percent accurate. In one test where particles were functionalized with a protein for detecting waterborne biological agents, 14 of the 16 particles tested showed the correct code, one was unreadable and one said there was no agent present when there was. Sailor said, however, that this is typical of first-generation experiments, and that his group will attempt to perfect performance over the next few years.
The smart dust produced by Sailor's group is the latest silicon-based technology developed to thwart terrorists. Earlier this year, Sailor announced silicon gunpowder a method of causing silicon chips to explode if a terrorist tries to access them with an incorrect security code. Sailor and his group have also developed a portable nerve gas detector (with a team headed by William Trogler, a professor of chemistry and biochemistry at the University of California, San Diego). With Trogler, Sailor has also devised a method of using tiny silicon wires in solution to detect molecules of TNT and picric acid, a common explosive used by terrorists. The Defense Advanced Research Projects Agency is supporting the work.