A collaboration of scientists in Europe has demonstrated what it believes is the first biochemical sensor based on an integrated slot waveguide. The intense, closely confined electric fields in the slot-waveguide microring resonator make the sensor at least twice as sensitive as biosensors that are based on traditional waveguides. Figure 1. The biosensor consisted of a microring evanescently coupled to a nearby bus (left). The slot was 200 nm wide, and the ring could be flooded with an analyte liquid (right). Images reprinted with permission of Optics Letters. Slot waveguides, developed several years ago by Michal Lipson’s group at Cornell University in Ithaca, N.Y. (Optics Letters, June 1, 2004), are different from conventional waveguides, in which light is guided in a high-index material by total internal reflection. In a slot waveguide, light is guided in a slot of low-index material only a few hundred nanometers wide, sandwiched between two strips of high-index material. The mode propagating in the slot wavelength is created by the interaction between eigenmodes in the two strips of high-index material. The slot-waveguide mode itself is a true eigenmode and is theoretically lossless. The intensity of light propagating in the slot-waveguide mode can be an order of magnitude greater than is possible in a conventional, total internal reflection waveguide. The scientists in Europe are associated with Universidad Politécnica de Madrid and with Universidad Politécnica de Valencia, both in Spain, and with Royal Institute of Technology in Stockholm, Sweden. They fabricated their microring sensor from Si3N4 on SiO2 (Figure 1). The top of the device was open to the air, so a drop of fluid could be added, flooding the ring. The fluid’s refractive index would change the effective (optical) circumference of the microring, a change that would show up as a shift in the ring’s resonant frequency. Experimentally, the scientists used several different concentrations of ethanol in deionized water to calibrate their sensor. They controlled the temperature of the device to 22 ±0.01 °C and probed it with light from a tunable laser, easily observing a significant shift in the ring’s resonant wavelength (Figure 2). These shifts were more than twice what would occur in a conventional microring resonator. They calculated that their sensor can detect a shift in refractive index as small as 2.3 × 10–4, limited by the 50-pm resolution of their tunable laser. Figure 2. The absorption peak of light propagating down the bus of Figure 1 — corresponding to the resonance peak of the microring resonator — shifted when different liquids were placed inside the waveguide. In this case, the liquids were various concentrations of ethanol in water, whoserefractive indices varied from 1.33 to 1.36. They also calculated the theoretical resonance-wavelength shifts for the cases when (1) the entire device is flooded with liquid, and (2) the slot remains dry while the rest of the device is flooded. Because there is significant electric-field strength outside the Si3N4 slabs that form the slot, there is still a wavelength shift when the slot itself remains dry but the fluid flooding the rest of the device changes. They compared these theoretical results with their experimental values (Figure 3). Figure 3. The experimental results were midway between the theoretical results for a filled and an empty slot. (Δηfluid = fluid’s change in refractive index, Δλ/λres = change in wavelength normalized to the unshifted (resonance) wavelength.) The experimental values were midway between the two extremes, so the scientists concluded that surface tension prevented the liquid from filling the tiny volume. A flow-injection system could fill the slot, further increasing the sensitivity of the biosensor. Optics Letters, Nov. 1, 2007, pp 3080-3082.