An extremely sensitive Raman sensor promises to provide new ways to detect a wide range of substances, from tell-tale signs of cancer to hidden explosives. Developed at Princeton University, the sensor relies on a completely new architecture and fabrication technique. It boosts faint signals generated by the scattering of laser light from a material placed on it, allowing the identification of various substances based on the reflected spectra. The sample could be as small as a single molecule. The technology is a major advance in a decades-long search to identify materials using Raman scattering, in which faint light signals reflecting off an object carry a signature of its molecular composition and structure. This schematic shows the architecture of a new Raman-based sensor developed at Princeton University. The chip is covered with structures that feature two key components: a cavity formed by metal on the top and at the base of each pillar; and metal particles of about 20 nm in diameter, known as plasmonic nanodots, on the pillar wall, with gaps of about 2 nm between the metal components. The small particles and gaps significantly boost the Raman signal. The cavities serve as antennae, trapping light from the laser so it passes the plasmonic nanodots multiple times to generate the Raman signal rather than only once. The cavities also enhance the outgoing Raman signal. (Images: Stephen Y. Chou) "Raman scattering has enormous potential in biological and chemical sensing, and could have many applications in industry, medicine, the military and other fields," said Stephen Y. Chou, the professor of electrical engineering who led the research team. "But current Raman sensors are so weak that their use has been very limited outside of research. We've developed a way to significantly enhance the signal over the entire sensor, and that could change the landscape of how Raman scattering can be used." Chou and his collaborators, Wen-Di Li, Fei Ding and Jonathan Hu, published a paper on their innovation in February in the journal Optics Express. The research was funded by DARPA. A micrograph of a sensor developed at Princeton for sensing Raman scattering shows the pillars that support metal components that gather light and amplify Raman signals. In Raman scattering, a beam of pure one-color light is focused on a target, but the reflected light from the object contains two extra colors of light. The frequencies of these extra colors are unique to the molecular makeup of the substance, providing a potentially powerful method to determine the identity of the substance, analogous to the way a fingerprint or DNA signature helps identify a person. In the 1970s, Raman signals were found to be much stronger if the substance to be identified is placed on a rough metal surface or on tiny particles of gold or silver. The technique, known as surface-enhanced Raman scattering (SERS), showed great promise, but even after several decades of research has proved difficult to put to practical use. The strong signals appeared only at a few random points on the sensor surface, making it difficult to predict where to measure the signal and resulting in a weak overall signal for such a sensor. This graph shows the boosted Raman signal (red) vs. an unenhanced signal (blue). The Princeton sensor uses surface-enhanced Raman scattering (SERS) to enhance the signals. By using a new sensor architecture for gathering and boosting the signals, the chip is a billion times more sensitive than without SERS enhancement. Abandoning the previous methods for designing and manufacturing the sensors, Chou and his colleagues developed a completely new SERS architecture: a chip studded with uniform rows of tiny pillars made of metals and insulators. One secret of the Chou team's design is that their pillar arrays are fundamentally different from those explored by other researchers. Their structure has two key components: a cavity formed by metal on the top and at the base of each pillar; and metal particles about 20 nm in diameter, known as plasmonic nanodots, on the pillar wall, with small gaps of about 2 nm between the metal components. The small particles and gaps significantly boost the Raman signal. The cavities serve as antennae, trapping light from the laser so it passes the plasmonic nanodots multiple times, rather than only once, to generate the Raman signal. The cavities also enhance the outgoing Raman signal. The team named its new sensor "disk-coupled dots-on-pillar antenna-array" or D2PA. So far, the chip is a billion times more sensitive than was possible without SERS boosting of Raman signals, and the sensor is uniformly sensitive, making it more reliable for use in sensing devices. Such sensitivity is several orders of magnitude higher than that previously reported. Already, researchers at the US Naval Research Laboratory are experimenting with a less sensitive chip to explore whether the military could use the technology for detecting chemicals, biological agents and explosives. In addition to being far more sensitive than its predecessors, the Princeton chip can be manufactured inexpensively at large sizes and in large quantities. Chou's team has produced these sensors on 4-in. wafers and can scale the fabrication to much larger wafer sizes. "This is a very powerful method to identify molecules," Chou said. "The combination of a sensor that enhances signals far beyond what was previously possible, that's uniform in its sensitivity and that's easy to mass produce could change the landscape of sensor technology and what's possible with sensing." For more information, visit: www.princeton.edu