Sensing with locally self-assembled 1D nanostructures
One-dimensional nanostructures have attracted considerable interest as potential building blocks and functional components in next-generation nanoscale sensing, electronics, and interconnect applications. They can be thought of as tiny wires, for example, with a nanoscale cross section and length generally measured in micrometers. A broad range of material systems may exhibit the geometrical characteristics of 1D nanostructures. For example, semiconducting nanowires, carbon nanotubes, and protein nanofibers can all be classified this way. In recent years, research efforts have focused on the synthesis and characterization of a wide range of these structures.1–7 With a wealth of well-characterized nanostructures and well-formulated processes to control material properties, demonstrations of nanoscale devices, particularly sensors, have become commonplace.1,8–13
Some of the most significant challenges to realizing 1D-nanostructure-based sensing, and particularly its large-scale adoption and production, are difficulties in nanostructure manipulation and integration into device architecture. Existing tools and techniques for fast, precise, and reliable manipulation lag in comparison to our ability to produce such structures. We must remember that a nanostructure never stands alone, and an interface to a larger-scale structure is required in any device configuration. Hence, we face a nanomanufacturing problem.
The production of 1D nanostructures is commonly realized using synthesis reactions, often referred to as bottom-up techniques. These methods involve growing wire-like nanostructures from atoms and molecules. A bottom-up technique is essentially a chemical reaction that occurs in the presence of particular reactants and under specific conditions. Relevant synthesis reactions generally require a catalyst, a vapor-phase reactant containing the desired nanostructure material, and a specific thermal environment.
Under appropriate conditions, these reactions allow for the rapid production of large quantities of 1D nanostructures. However, post-synthesis processing and integration are required to transform them into useful sensing elements. Harvesting, tedious manipulation of individual nanostructures into place, and sequential contact formation may all be required to yield a functional device architecture. These steps increase the cost and processing complexity. Additionally, the high temperatures required for most reactions rule out synthesis on substrates with temperature-sensitive materials or devices. Thus, the ability of most high-quality nanostructures to be directly synthesized and simultaneously integrated with other key device components is limited. By confining the location of the synthesis reaction, however, we can use these bottom-up techniques for site-specific synthesis and directed self-assembly. The spatial extent of the reaction can be controlled by locally eliminating one of the conditions required for it to take place.
The confinement of the required thermal environment, for example, permits localization of the synthesis reaction. Our group first demonstrated the feasibility of this approach with site-specific synthesis of silicon nanowires and carbon nanotubes on suspended silicon microstructures.14 The microstructures were locally heated by resistive heating. We deposited the catalyst required for synthesis chip-wide, and introduced the vapor-phase reactant into a room-temperature reaction chamber. Microstructure design determines the power input needed to realize suitable heating levels for the reaction to take place. In practice, the experimental setup requires a feedthrough into the synthesis chamber that provides a connection to an external power supply. In a more recent demonstration by Luo and others, induction heating was coupled with the bottom-up approach to realize localized synthesis of zinc oxide (ZnO) nanowires.15,16 In this case, local vapor transport and extremely rapid synthesis rates characterize the process.
These approaches offer opportunities for the self-assembly of 1D nanostructures and their integration into systems and devices. The microscale design for this process calls for adjacent yet thermally and electrically isolated silicon microstructures, where localized 1D nanostructure synthesis eventually bridges the gap between the two microstructures, yielding a self-assembled two-terminal micro-nano system. The synthesis, self-assembly, and integration processes all take place within a single processing step. The distance between the bulk structures defines the length of the nanostructures, and the application of a localized electric field effectively assists their alignment and orientation.17 The bulk microscale structures serve as the means for interfacing with the nanostructures and essentially become electrodes.
This particular approach yields a device architecture with suspended nanostructures that is ideal for sensing applications, as it makes the entire surface area of the nanostructure available for sensing interactions. Thus, an enhancement in sensing capabilities is predicted compared to surface-bound 1D-nanostructure-based devices. The processing sequence needed to realize the sensing architecture and implementation is shown in Figure 1.
Using localized synthesis and self-assembly, our group has demonstrated proof-of-concept gas, pressure, and UV sensing applications.15,16,18–21 The self-assembled systems are ready for sensor testing as soon as synthesis and integration are complete, as Figure 1(e) shows. This is a significant advantage of this approach. We have examined hydrogen sensing using palladium-coated silicon nanowires. We recorded the change in resistance across the nanowires upon hydrogen dissociation onto their surfaces and correlated it to the presence of hydrogen.18 Our group has also shown that using the electrothermal effect in sensing with multiwalled carbon nanotubes (MWCNTs), both gas pressure and gas species can be detected.20,21 Upon exposure to gaseous environments, we found the resistance of a heated MWCNT to change following conductive heat-transfer variance of the gas molecules. Pressure sensitivity was correlated to the input voltage. Input voltages as low as 1V yielded good sensitivity over a wide pressure range.21 The low power requirement is a notable advantage of this approach. We pursued oxygen sensing using a ZnO nanowire platform. The sensing mechanism was based on a resistance change across the nanowires, as adsorbed oxygen molecules capture free electrons in ZnO nanowires.16 The sensitivity of these sensors improved under UV illumination, which may facilitate oxygen desorption. Preliminary data suggests sensitivity improvements on the order of 500%.
To conclude, we have demonstrated a range of 1D-nanostructure-based sensing applications using localized synthesis and self-assembly techniques. The ease with which self-assembled systems can be turned into functional sensing devices shows the remarkable potential of these integration and assembly methods. Further, we illustrate that sensing is possible with a range of nanostructured materials and is applicable to various sensing applications. The advantages of our approach also include compactness, low power consumption, and complete integration with on-chip microelectronics. Sensor performance is influenced by a number of factors. We are most concerned with resistance at the nanostructure/silicon-electrode interface and nanostructure quality in terms of crystal structure and the presence of defects arising during synthesis, self-assembly, and integration. These issues are the subject of ongoing studies. We believe that our techniques can be extended to other nanostructured materials and additional sensing applications. In future applications, we are interested in integrating biological 1D nanostructures that synthesize and self-assemble under even more favorable conditions and may self-heal and regenerate following failure.
The author thanks Liwei Lin and members of his group at the University of California, Berkeley, for their contributions to this effort over the years.
