Exploiting an accidental discovery, researchers at California Institute of Technology in Pasadena and at Rensselaer Polytechnic Institute (RPI) in Troy, N.Y., have demonstrated a method to produce bright color fields over large surfaces, creating any color desired in areas as small as 100 μm2. The technique, which depends upon optical interference from arrays of submicron dimples, could be used to make displays, lab-on-a-chip liquid flow monitors and index-of-refraction change sensors.Color changes resulting from fluid changes can be used to read out a microfluidic sensor or could form the basis for a display. A submicron dimple array milled into silicon changes color from blue to orange when air is replaced by water (b, c and d). A schematic of the demonstration device is shown in (a). The color change is simultaneous for all dimples across the column. Images courtesy of J. Jay McMahon, Rensselaer Polytechnic Institute.“Ultimately, we would like to be able to make disposable sensors for a variety of chemicals. We hope to build large displays with highly readable output in ambient light,” said J. Jay McMahon, an engineer at the Center for Integrated Electronics at RPI.Team member Henri J. Lezec, now at Caltech, first observed the phenomenon while doing research at Centre National de la Recherche Scientifique in Strasbourg, France. Using focused ion beam milling, he manufactured arrays of dimples in standard silicon in which the constructs’ diameter, pitch and depth varied over a large range. Surprisingly, many of the arrays exhibited vivid colors when observed under a microscope. Furthermore, adding a drop of methanol to the surface brought about an instantaneous and dramatic color change. Dimples milled into reflective silicon at different depths give rise to different colors when viewed under reflection. Here, an array of increasing depth but with constant spacing shows this effect. Dimples in Field 1 are 96 nm deep (b); those in Field 12 are 448 nm deep (c), and those in Field 23 are 800 nm deep (d). Observed color changes result from interference. The effect arises because the light reflected from the bottom of the dimple interferes with that from the surrounding surface. For this to happen, the dimples must be more than half a wavelength in diameter, which allows light to propagate up and down the cylinder. In addition, the dimple spacing must be approximately equal to one wavelength, allowing the two interfering components to be about the same intensity. Adding a liquid alters the interference and, therefore, the color.The discovery owes something to the capabilities of focused ion beam milling, which enables the creation of dimples of varying depths and spacings in a single step.Dimple depth increases from 224 nm on the left to 736 nm on the right. The observed color in air (a) with a refractive index of 1.0 changes when methanol (refractive index 1.3) is introduced. The change is instantaneous and simultaneous for all dimples.“A more traditional approach, combining standard optical or electron-beam lithography with wet or dry etching, would be incredibly tedious, involving one lithography step per depth,” McMahon said.The team etched dimples into polished silicon wafers, demonstrating a depth-dependent reflection spectrum. They also placed a microchannel connecting two wells at either end atop a line of dimples of uniform depth. They filled one well with air and the other with water. Changing the pressure switched the color of the dimples instantaneously from blue to orange as water displaced air.That color change could be the basis for both displays and sensors. The color could be switched on command or in response to the presence of a chemical, without the use of polarizers, liquid crystals or color filters. That makes the method simpler than other approaches.Achieving a display or sensor will require more research and, perhaps, a change in the fabrication scheme. Focused ion beam milling probably is too slow for mass manufacturing. High-volume production may require stamping or embossing to create the dimples, and the researchers are investigating ways to do this, according to McMahon. “We are looking into nanoimprint lithography for high-volume applications.”Nano Letters, ASAP edition, Dec. 23, 2006.