A new class of biosensors could be at hand, according to Northwestern University researchers. They constructed composite metal-dielectric-metal nanoantennas on the facet of a quantum cascade laser, thereby boosting the incident optical field intensity in the mid-IR as much as 4000-fold. Their device had a lasing mode spot size more than 10 times smaller than the laser wavelength. This light squeeze solves a fundamental problem, said team leader Hooman Mohseni. The wavelength of a mid-IR laser is thousands of times larger than the size of biomolecules, making interaction with the light poor, signals weak and detection difficult. Metal-dielectric-metal nanoantennas (scanning electron microscope image on right) fabricated on and integrated with the facet of a quantum cascade laser (center) concentrate infrared emission (top left inset). The laser’s current voltage characteristics are shown. Reprinted from Optics Letters. Images courtesy of Dibyendu Dey and Hooman Mohseni, Northwestern University. This obstacle can be overcome if the researchers can reduce the detection volume by up to an order of magnitude, according to Mohseni, an associate professor of electrical engineering and computer science. He added that quantum cascade lasers are a natural choice for a source in the range of 2 to 30 μm, a region where many important biomolecules have an identifying spectral signature. Metal nanoantennas have been used in the past to concentrate laser light in this range. These antennas interact with the light and compress it into a small volume. Side view of simulated intensity profile for metal dielectric metal nanoantenna (Au/SiO2/Au, 70/30/70 nm) at a resonance antenna length of 2 μm. Reprinted from Optics Letters. The Northwestern group used a different type of nanoantenna, fabricating the devices out of multiple layers of metal and dielectric. They chose this approach, said graduate student Dibyendu Dey, because these composite-material nanoantennas increase near-field enhancement. Indeed, the group achieved four times the enhancement previously seen, according to an Aug. 15, 2010, Optics Letters paper, of which Dey was the lead author. Thickness matters The scientists simulated the effect of the thickness of a dielectric layer of silicon dioxide from 0 to 100 nm. In these calculations, they kept the total thickness of the nanoantennas fixed at 170 nm. Thus, those devices with thicker oxide had thinner layers of gold sandwiching the dielectric. The metal thickness ranged from a high of 85 to a low of 35 nm for each layer. The researchers found that a 30-nm oxide thickness delivered the best results. They then constructed the optimum composite nanoantennas on the facet of a quantum cascade laser by first depositing the alternating gold and oxide layers on top of a 100-nm buffer oxide. Using a focused ion beam, they then carved coupled nanorods out of the deposited layers, with each rod about 500 nm wide and about 2 μm long. They fabricated the two so that their short ends faced each other with a separation in the tens of nanometers. The antenna had a resonant wavelength of about 2 μm. To evaluate the results, they turned to a custom-built near-field scanning microscope. With this, they measured the optical hot spots, finding that these results agreed well with what simulations had predicted. The lasing spot size was about 450 nm, an order of magnitude smaller than the laser wavelength. As for the future, quantum cascade lasers are already being used for gas sensors. Because it concentrates light, the nano-antenna approach promises to enable quantum cascade laser-based biosensors. Since the nanoantennas are integrated onto the laser itself, the resulting device will be very compact. As an added bonus, device operation can be tuned by varying laser power or by employing different layer combinations. Once the remaining problems are solved, the researchers are certain that there will be considerable commercial interest in the technique. One challenge is how to position the biomolecules in the optical hot spot, which is at specific locations with respect to the nanorods. The researchers are evaluating how best to achieve this, Dey said. “We are working on basically two schemes: to use dip-pen nanolithography to specifically position the molecule at a particular location and a microfludic approach.”