Precision machining produces tiny light-guiding cubes to advance information technology
Drilling with the beam of an electron microscope, scientists at the Department of Energy’s Oak Ridge National Laboratory have precisely machined tiny electrically conductive cubes that can interact with light and organized them into patterned structures that confine and relay the electromagnetic signal of light. This demonstration is a step towards potentially faster computer chips and more perceptive sensors.
The apparent magic of these structures comes from the ability of their surfaces to support collective waves of electrons, called plasmons, with the same frequency as light waves but with much tighter confinement. Light-guiding structures are measured in nanometers, or billionths of a meter – 100,000 times thinner than a human hair.
“These nanoscale cube systems allow for extreme confinement of light to specific locations and tunable control of its energy,” said ORNL’s Kevin Roccapriore, first author of a study published in the journal Small. “It’s a way to connect signals with very different length scales.”
The feat may prove critical for quantum and optical computing. Quantum computers encode information with quantum bits, or qubits, determined by a quantum state of a particle, such as its spin. Qubits can store many values compared to the single value stored by a typical bit.
Light – electromagnetic radiation that propagates by massless elementary particles called photons – replaces electrons as messengers in optical computers. Because photons travel faster than electrons and do not generate heat, optical computers could have higher performance and power efficiency than conventional computers.
Future technologies could use the best of both worlds.
“Light is the preferred way to communicate with qubits, but you can’t connect contacts to it directly,” said lead author Sergei Kalinin of ORNL. “The problem with visible light is that its wavelengths range from about 380 nanometers for violet to about 700 nanometers for red. That’s too big because we want to make devices only a few nanometers in size. This work aims to create a framework for advancing technology beyond Moore’s Law and classical electronics. If you’re trying to put ‘light’ and ‘small’ together, that’s exactly where plasmonics comes in.”
And while there is a great future in plasmonics, the ORNL-led realization may help overcome a signal size mismatch that threatens the integration of components made of different materials. These hybrid components will have to “talk to each other” in next-generation optoelectronic devices. Plasmonics can fill the gap.
Plasmonic phenomena were first observed in metals, which are conductors due to their free electrons. The ORNL team used cubes made of a transparent semiconductor that behaves like a metal, indium oxide doped with tin and fluorine.
The fact that the cube is a semiconductor is the key to its energy tunability. The energy of a light wave is related to its frequency. The higher the frequency, the shorter the wavelength. Visible light wavelengths appear to the human eye as colors. Because a semiconductor can be doped – that is, a small impurity can be added – its wavelength can be shifted across the spectrum.
The cubes in the study were each 10 nanometers wide, which is much smaller than the wavelength of visible light. Synthesized at the University of Texas at Austin by Shin-Hum Cho and Delia Milliron, the cubes were placed in detergent to prevent clumping and pipetted onto a substrate, where they self-assembled into a two-dimensional network. A shell of detergent surrounded each cube, spacing them evenly. Once the detergent was removed, the matrices were sent to ORNL.
“The fact that the cubes don’t directly touch each other is important for collective behavior,” said Roccapriore, who organized the cubes into various structures. “Each cube individually has its own plasmon behavior. When we put them together in geometries like a nanowire, they talk to each other and produce new effects that are not typically seen in similar geometries that aren’t made up of individual elements. “
The study builds on previous work to sculpt three-dimensional structures as small as one nanometer with an electron beam. “The current paper proves that the plasmonic effect, as well as the structure, can be sculpted,” Roccapriore said. “Ultimately, we are interested in the electron wave – where is it and what is its energy? We control both of those things.”
Kalinin added, “We want to move from using what exists in nature by chance to making materials with the right answers. We can take a system of cubes, shine light on them, and channel the energy into small localized volumes exactly where we want them to be. .”
The project was natural for Roccapriore, who did a lot of electron beam lithography in graduate school and even built a machine in his garage to fabricate and machine 3D printed structures. At ORNL, experimenting with the beam of an electron microscope, he adjusted his current to intentionally switch from imaging mode to modification mode. He discovered that he could remove pieces of cubes or whole cubes from an array to create patterned objects at will. He also discovered that just as the addition of chemical elements allows the tuning of cubic energies, so does the selective removal of chemical elements. Such atomic precision is possible with scanning transmission electron microscopy, or STEM.
The key to characterizing plasmonic behavior in individual cubes and among collective cube assemblies was a technique called electron energy loss spectroscopy. It uses a STEM instrument with an electron beam filtered at energies in a narrow range. The beam loses energy as its electrons pass through the sample, interact with electrons in the material, and transfer some energy to the system by exciting the plasmons.
Electron energy loss spectroscopy provides in-depth insights into exotic physics and quantum phenomena related to plasmonic behavior,” said co-author Andrew Lupini of ORNL, who has helped map electron energies in cubes and arrays of cubes. Lupini is one of the aberration-corrected STEM developers who made pioneering advancements possible. “Electronic energy loss spectroscopy allows us to analyze the evolution of plasmonic responses in real time as the cubes are sculpted. We can understand the relationships between the arrangements of cubes and their plasmonic properties.”
Scientists plan to create a library of relationships between materials, structures and plasmonic properties. This new knowledge will provide the fundamental understanding needed to eventually mass-produce structures capable of directing the flow of light in plasmonic nanocircuits. According to Roccapriore, “the idea is to understand relationships using machine learning and then automate the process.”