Nanoneedles Point the Way to Sharper Sensors
Last year, Connie Chang-Hasnain and graduate student researcher Linus Chuang were searching for a better lab recipe for growing nanowires, conductive threads so thin that every atom they contain has a significant effect on their overall electrical properties. Following the vapor-liquid-solid (VLS) technique for creating semiconductor crystal nanowires, they deposited successive layers of gallium arsenide onto a silicon wafer substrate. But in one low-temperature batch, an area of the silicon lacked the usual gold nanoparticles from which each crystal grows.
Under careful examination of the region, they didn't find what they were expecting. Instead of uniform-diameter threads sticking up, they saw tall, needle-like pyramids with hexagonal bases and sharp points. They had discovered a new nanostructure.
"It's a single crystal that grows at an ultra-sharp six- to nine-degree angle," Chang-Hasnain explains. "Its tip may be the sharpest point we’ve ever seen." Chang-Hasnain (M.S.’84, Ph.D.’87 EECS) is the John R. Whinnery Professor in Electrical Engineering and Computer Sciences and director of UC Berkeley’s Center for Optoelectronic Nanostructured Semiconductor Technologies. She also chairs the Nanoscale Science and Engineering Graduate Group.
Gallium arsenide is classified as a group 3-5 semiconductor because it combines elements from the group 3 and group 5 columns on the periodic table. Chang-Hasnain found that other 3-5 compounds, such as aluminum gallium arsenide, also formed the structures. And by alternating and selectively etching successive deposits of different 3-5 semiconductors, she confirmed the needles' internal composition and how they grow up from the surface. The key to their formation, she found, is that the atoms in the silicon substrate, which is also a crystalline structure, are spaced 4 percent closer together than the atoms in the crystal that grows on top of it. The 3-5 atoms in the cultivated substance self-assemble into 3-D clusters around the mismatch with the lattice underneath, and this is what starts the needle growth.
As with flat semiconductor materials, you can tune the chemistry of the needles' different layers to turn them into solid-state electronic components. Using this method, for example, Chang-Hasnain has created nanoneedle diodes and resistors. But some needle formulations work in a more specialized way as avalanche photodiodes, an ultrasensitive type of light detector.
Avalanche detectors are expensive components used in long-range optical communications and medical imaging. They depend on an external electric field to amplify the number of electron-hole pairs formed when a semiconductor surface absorbs a photon. But the nanoneedle's atomic arrangement naturally forms its own electric field from tip to base; when a photon hits the tip, this causes the same avalanche effect downward, possibly even at a higher amplification and without the need for as much external power. "The needle acts like a little vacuum cleaner, drawing the hole up and the electrons down, resulting in an efficient current path," Chang-Hasnain says.
The needles can also be operated in reverse, to convert voltage into photons. And, by changing the composition of the needles, you can tune the wavelength of light they emit or detect. "What we really want to do is make a laser out of these," Chang-Hasnain explains, "a CMOS-compatible, low-temperature laser on silicon. So far, it looks like the needles have all the properties we would need to do this."
Such photodetectors and lasers could act as nanoscale, nanopower–consuming bridges between electrical and optical signal, which would revolutionize microprocessor design. For decades, shrinking transistors have made chips faster and more powerful, but as the number of transistors has grown, so has the number of interconnections between them. In today's chips, shuttling electrical signal from layer to layer and gate to gate takes more time, consumes more energy and generates more waste heat than running the logic gates themselves. Nanoneedle optics would enable radically different chip designs, porous chips that would replace the dense spaghetti of hard interconnections with light channels between layers, speeding and simplifying signal routing the way a single optical fiber can upgrade a bulky wire cable in a network.
The needles have other potentially revolutionary applications, such as solar cells and other light energy–capturing surfaces. Avalanche nanoneedles are far more efficient than conventional solar cells at converting light to voltage, provided the photons hit their tips. Perhaps a dense forest of nanoneedles or a fabric of needles and reflectors could convert a high percentage of incident light, that is, the direct light that falls onto a surface, into usable voltage underneath.
But wait—there's more. Not only do nanoneedles capture photons (particles of light energy); they also capture and propagate phonons, particles of vibration (or sound) energy. And because molecules vibrate in ways characteristic of their structure and composition, the needles could potentially facilitate ultra-sensitive chemical sensors that analyze the vibrations using a technique called surface-enhanced Raman spectroscopy. "When a molecule sits on a needle coated with a thin layer of metal, the needle acts like an antenna," Chang-Hasnain explains. "You can detect the vibrations on the metal and identify the molecule."
Chang-Hasnain summarizes, “The nanoneedle's single crystalline formation makes it the most stable form of sharp tip. These structures are so new that we are finding something new about them every day. It's exciting; we are just beginning to see what they can do.”