Sculpting a silicon wafer surface one atom at a time
Shaping the topography of a silicon wafer surface , atom-by-atom, requires precise manipulation of matter at the atomic scale and is critical for the miniaturization of electronic devices. Researchers from China and the United States have developed a simple method to etch silicon surfaces based on chemical reactions that occur as a result of mechanical stresses. The process has reduced energy demands and makes zero use of dangerous chemicals. The study, published in a recent issue of Nature Communications, advances research toward nanofabrication with ultimate precision, opening new opportunities for advanced nanoelectronics with new functionalities.
Etching-based lithography, which uses wet chemicals or high-energy plasma, is a common nanomanufacturing method for silicon; however, controlling the reaction kinetics with these processes has proven difficult beyond the nanoscale, down to the single atomic level.
A team of researchers from Southwest Jiaotong University, Tsinghua University, and The Pennsylvania State University have demonstrated the removal of silicon atoms from the surface of a single crystalline silicon wafer by using the less common mechanochemical reactions to produce a controlled pattern without damaging the crystalline lattice beneath the processed area.
A reaction is considered mechanochemical when breakage of primary bonds is induced or promoted by mechanical forces. The type of mechanochemical reactions used in this study are tribochemical reactions (“tribo” meaning rubbing or friction in Greek) and they occur as a result of mechanical shear at sliding interfaces. “We utilized the fact that such chemical reactions can occur only at the topmost surface, exposed to the gas phase through which other reactants (like water) are supplied. Thus, we were able to achieve the ultimate precision in the material removal process—one single atomic layer,” says Seong H. Kim, a professor at The Pennsylvania State University.
The first step in this mask-free and chemical-free lithographic process is treatment of the silicon wafer to produce a hydrophobic surface passivated with hydrogen atoms. Scanning probe microscopy was then used to sculpt the surface at specific locations as well as for the in situ topography scanning.
At the tip of the scanning probe, attached to a cantilever, a silicon dioxide (or silica, SiO2) microsphere with a diameter of 2.5 μm rubs the surface of the wafer in a controlled humid environment. The presence of water at the interface between the microscope tip and the wafer surface is critical, as it significantly decreases the critical energy barrier for the mechanochemical reactions.
The mechanochemical reactions promoted by the sliding of the SiO2 tip against the silicon substrate in the presence of water molecules were depicted by molecular dynamics simulations as a three-stage process. First surface atoms react with water molecules to generate surface hydroxyl species. Then interfacial bridge bonds form as a result of a dehydration reaction between two surface hydroxyl groups across the interface. Third, the mechanical shear action causes the dissociation of the substrate bonds and leads to the removal of a Si atom from the substrate.
According to Kim, the team’s biggest effort was to find the right process conditions (humidity and applied load ranges). “Once we found the initial conditions it was relatively straightforward, although not easy, to design and execute control experiments and simulations to figure out the governing mechanism,” he says.
“This study represents a significant step toward development of robust nanomanufacturing methods at the atomic scale,” says Ashlie Martini, a professor of mechanical engineering at the University of California–Merced, who was not involved in the study. “It not only demonstrates controlled, single-layer atomic removal with highly relevant materials, but combines simulations and experiments to reveal the reaction pathways underlying the removal process. This encourages exploration of other opportunities to use shear in addition to or instead of traditional drivers (heat, charge, light, etc.) to alter the rates and pathways of chemical reactions,” she adds.
Kimani C. Toussaint, a professor at the University of Illinois at Urbana-Champaign, who was also not involved in the study, says, “Selective patterning of silicon has been critical to the electronics industry. The authors have come up with a clever way to pattern crystalline silicon with atomic-depth resolution without the use of masks or harmful chemical etchants, and without the need of a vacuum chamber.” Toussaint believes that the process is very likely to be further explored by the nanomanufacturing community.
“Since the technique uses a scanning probe tip, a challenge will be to think of ways to scale up this process to achieve high volume nanomanufacturing of silicon,” Toussaint says. For Kim, scaling-up does not mean a larger volume, but “numbering-up” via simultaneous processing of multiple patterns on a large area of a silicon wafer.
Many more challenges lie ahead for Kim and his colleagues before the technique can be used widely. “At this moment, the process takes place at one specific location at a time. The hope is to combine the reaction with a process using multiple scanning probes (such as IBM’s Millipede project),” he says. The researchers are also testing silicon wafers with different crystallographic orientations and they are planning to expand the range of wafer materials beyond silicon. They are also looking into combining other chemical reactions to functionalize the mechanochemically etched regions for chemical sensing or to use them as templates for more complicated functional modality.
Read the article in Nature Communications.