Arxiv – Six degree of freedom atomic-scale manipulation using carbon nanotube bundles The designs can support the 9 tooltips designed by Freitas and Merkle.
Abstract. Scanning probe imaging and manipulation of matter presents is of crucial importance for nanoscale science and technology. However, its resolution and ability to manipulate matter at the atomic scale is limited by the rather poor control over the fine structure of the probe. In the present communication, a strategy is proposed to construct a molecular nanomanipulator from ultrathin single-walled carbon nanotubes. Covalent modification of a nanotube cap at predetermined atomic sites makes the nanotube act as a support for a functional ―tool-tip‖ molecule. Then, a small bundle of nanotubes (3 or 4) with aligned ends can act as an extremely high aspect ratio parallel nanomanipulator for a suspended molecule, where protraction or retraction of individual nanotubes results in a controlled tilting of the tool-tip in two dimensions. Together with the usual SPM three degrees of freedom and augmented with rotation of the system as a whole, the design offers six degrees of freedom for imaging and manipulation of matter with precision and freedom so much needed in modern nanotechnology. A similar design might be possible to implement with other high-aspect ratio nanostructures, such as oxide nanowires.
The present communication describes a class of nanoscale parallel manipulators based on carbon nanotube bundles. The manipulators offer precise control over the position and orientation of individual molecules, thanks to the well-defined structure of constituent nanotubes and to the two additional degrees of freedom that such systems provide, compared to regular scanning probes. An important step is the choice of carbon nanotube types so that their tips can be functionalized at predictable atomic sites.
The functional molecules can then be attached by either strong covalent C-C bonds or reversible dative bonds between substitutional B and N atoms in the parts of the assembly. The designs have been demonstrated to be thermodynamically feasible, and pathways to their practical fabrication using present-day or nearest future technology have been suggested. Although manipulators such as those described above can be expected to substantially improve the spatial resolution of scanning probe microscopy, the true diversity of potential applications comes from the various kinds of functional molecules that they can support; even without individual nanotube actuation, tight locking of the molecules will allow improved control over their position and orientation. Here, designs that can support all 9 tooltips from the minimal toolset for positionally controlled mechanosynthesis have been provided. If built, they could serve as stepping stones from current scanning probe technology towards more efficient autonomous positioning systems required for high-throughput deterministic manipulation of matter at the atomic scale, ultimately leading to the much anticipated prospects of machine-phase diamond and graphitic nanotechnology. Less remote applications of present and similar parallel nanomanipulator systems will include atomic force microscopy of complex biological objects such as enzymes with improved resolution and chemical sensitivity (including STM-level precision in liquid media), as well as controlled assembly of nanostructures from nucleotides, amino acids and other molecular building block types.
Before the implications of the above designs can be discussed, possible strategies of fabricating the proposed structures have to be reviewed. This includes synthesizing the required components and assembling them into a working structure.
As of present, carbon nanotube probes are typically grown in situ on SPM tips using some variation of chemical vapor deposition technique, with the possibility of even wafer-scale fabrication. However, CVD-grown nanotubes typically have diameters > 1 nm, meaning that they are possibly too thick for our purposes, and their specific type is hard to control. On the other hand, ultrathin SWCNTs (down to 0.4 nm diameter) can be selectively grown within zeolite pores, or inside larger diameter CNTs with the possibility of controlling the type of as-grown nanotube by the choice of catalyst type and external conditions. The inner tube could subsequently be extracted from the resulting double-wall nanotube by mechanical means (so-called ―sword-in-sheath‖ failure of the outer wall) or using electrical current heating. Even if the nanotubes are grown uncapped, it should nevertheless be more or less straightforward to close their ends; on-demand capping of carbon nanotubes has previously been demonstrated, at least, for multiwall carbon nanotubes.
Given all the difficulties of fabrication and processing of ultrathin carbon nanotubes, it might be desirable to use nanocones or conically-terminated multiwall nanotubes, since these structures can have very sharp tips with clusters of pentagons. Although chemical modification of nanocones is much less studied compared to nanotubes, quantum chemical calculations suggest that functionalization of nanocones should occur predominantly at the tip, offering at least some spatial control over functionalization. Finally, it should be noted that perfect control over the functionalization site is not an absolute necessity: techniques such as field emission measurements with a second movable probe could in principle be utilized to determine functional group position after the functionalization has been carried out, thus enabling the use of other carbon nanostructures besides the (6, 0) nanotube.
Individual as-grown carbon nanotubes could then be transferred onto separate actuators, and their free ends joined together to form a self-supporting bundle. This would be most easily achievable if all actuators had three degrees of freedom, but one degree per actuator should in principle be sufficient, provided the tubes are long enough (or the actuators close enough) so that they can be joined using an additional 3-dof manipulator. After the bundle has been formed, DNA hairpins could be used to make a ―knot‖ clipping the bundle together and allowing individual nanotubes to be routed to their independent actuators, although it is quite probable that simply relying on the mutual attraction of carbon nanotubes would be sufficient. Any excess length of the nanotubes could be trimmed in situ with, e.g., an electron beam.
Covalent functionalization of carbon nanotubes is a well-established technique, and no insurmountable obstacles are to be expected from this side. Similarly, organic synthesis methods are more than capable of producing molecules with appropriate linkers attached, as long as a desired functional molecule has been chosen. The possibility of successful synthesis of the particular tooltip molecules shown in figures above has already been discussed in the corresponding references.
Given the freedom to choose the functional groups on both sides—nanotube tips and the molecules—it appears that the rest (putting the functionalized molecule on a pre-assembled functionalized nanotube bundle) is also within the reach of present-day scanning probe manipulation technology.