The nanobliss artifacts are small intricate structures that are fabricated from organic and inorganic materials such as carbon and silicon. The structures range in size from far invisible to nearly visible to the naked eye, but often appear to be scale-independent and are reminiscent of much larger objects in everyday life and landscape. Techniques of materials science, chemistry, micro-fabrication, self-assembly, mechanical shaping, scanning electron microscopy, optical photography, and digital image processing are used to create the structures and images shown here.

Carbon Nanotubes
Nanobliss is largely based on technologies for synthesis of carbon nanotube structures, which were developed by John Hart through his research at the Massachusetts Institute of Technology. Carbon nanotubes (CNTs) are cylindrical molecules of carbon atoms, where the carbon atoms are arranged in a hexagonal lattice as in graphite (Figure 1). Because carbon-carbon bonds are very stable and strong, and because CNTs are seamless and have a very small diameter (1-100 nanometers, or 0.00000004-0.000004 inches), CNTs have exceptional properties. High-quality CNTs have several times the strength of steel piano wire at one-fourth the density, at least five times the thermal conductivity of copper, and very high electrical conductivity and current-carrying capacity. These properties have generated broad interest in CNTs, for potential applications such as next-generation electronics where individual CNTs are transistors, to advanced composites where trillions of CNTs work together to form the structure of an airplane wing. For example, a fully-loaded 747 (~400 tons) could hang from a 5 millimeter (1/4 inch) diameter rope made from continuous parallel CNTs!
Figure 1. The structure of a CNT: a seamless cylinder of carbon atoms arranged in a hexagonal lattice. The schematic at left is of a single-wall CNT, and the schematic at right is of a multi-wall CNT.
CNTs are made by the high-temperature process of chemical vapor deposition (CVD). A carbon-containing gas is converted in to CNTs using small catalyst particles which act like "seeds" for growth; wherever a catalyst particle is placed, a CNT starts growing when the catalyst is heated to the necessary reaction temperature and the carbon gas is introduced (Figure 2). Many further details of the growth process are described in John's Ph.D. Thesis [1] and related publications; download an extended abstract, or contact John to obtain his complete thesis.
Figure 2. Schematic of growth of a CNT from a metal catalyst nanoparticle on a substrate.
Fabrication Techniques
For Nanobliss, CNTs are grown on substrates such as silicon wafers: first, the catalyst is arranged on the surface of the wafer, and then the carbon source is introduced to grow the nanotubes by placing the catalyst-coated substrate in a sealed reaction chamber. When a relatively high density of catalyst particles is placed on the substrate, the CNTs align with each other and grow perpendicular to and upward from the substrate, to form a CNT "forest". By controlling the density of the catalyst particles as well as the reaction temperature and chemistry, we can grow these CNT forests to millimeter heights [2].
A hierarchy of length scales is involved (Figure 3). At the molecular (nanometer) scale, carbon atoms organize at a catalyst particle to produce a single CNT. The individual graphitic walls of each CNT are visible using a transmission electron microscope (TEM), as each CNT is approximately 10 nanometers in diameter. At the micrometer scale, the CNTs are self-organized to form the forest in which the CNTs are roughly parallel and aligned; this is seen using a scanning electron microscope (SEM). At the millimeter scale, the microscopic shape of the forest, in this case the sidewall, is seen to stretch to a height of millimeters, and is again imaged using a scanning electron microscope. At the centimer scale, we see an optical photograph of a thick CNT forest on a substrate, resting on a human fingertip. There are approximately 20 billion CNTs on this substrate, and each CNT has an aspect ratio (length/diameter) of approximately 400,000:1.
Figure 3. Ascending hierarchy of length scales in assembly of a CNT forest (click on image to enlarge): (1) TEM image showing concentric layers of a multi-wall CNT; (2) SEM image showing alignment among CNTs in sidewall of a forest; (3) SEM image of sidewall of a 2 mm tall forest; (4) optical image of forest on a silicon substrate (approximately 1 cm x 1 cm) resting on a human fingertip.
Chemical, mechanical, and thermal assembly processes convolve to realize the CNT structures which are imaged for the Nanobliss galleries. In the simplest case, the catalyst is deposited in a uniform layer, and a forest of CNTs (as shown in Figure 3) which entirely covers the top surface of the substrate. In particular cases, differential growth rates across large numbers of CNTs cause the CNTs to bend or wrinkle. Mechanical friction among the growing CNTs can cause groups of CNTs to break away from the forest or to grow taller than neighboring groups. Additional complexity is added by patterning the catalyst, such as by photolithography, so CNTs grow only in certain areas on the substrate. This produces arrays of microstructures and complex shapes (Figure 4). By introducing spatial and temporal gradients in the reaction conditions or by spatially varying the size of the patterns, the shapes can be influenced to grow to different heights or to lean in particular directions. The gallery of self-organized and patterned architectures exemplifies these techniques. In the gallery of logos and popular impressions, structures are engineered to represent exemplary logos and other advertising symbols. In principle, any two-dimensional drawing can be replicated in a three-dimensional CNT structure by patterning the catalyst.
Figure 4. Fabricating a CNT microstructure (click on images to enlarge): (left) by patterning of the catalyst on the substrate; (right) "molding" CNTs into three-dimensional structures by applying mechanical pressure and confinement during growth
A further technique uses mechanical pressure to define the shapes of the CNT structures; in research, we discovered that CNT growth can output a significant force, and therefore forces can be used to affect the growth process [3]. For example, confining the CNTs inside a mold causes growth to take the shape of the mold (Figure 4). In comparison to patterning of the catalyst which defines a two-dimensional template for growth, using a mold defines a three-dimensional template [4]. Pressing on the CNTs also causes them to bend, and by controlling the applied pressure we can cause the CNTs within a forest to be "wavy" rather than aligned. Examples are in the gallery of mechanically-shaped structures.
This page describes just a few of many current and possible fabrication techniques for Nanobliss. The catalyst particles can be organized using polymer chemistry to precisely control the density of CNTs on a substrate [5]. We can watch growth by placing the substrate on a locally-heated platform in an open-view reaction chamber [6]. By resistively-heating a silicon platform beyond its melting temperature, followed by sudden re-solidification, we create intricately branched silicon structures.
Imaging
Because the CNT and silicon structures are electrically conductive, they can be imaged using a conventional scanning electron microscope (SEM). Compared to optical imaging where interaction of light with the subject (sample) forms the image, in electron microscopy interaction of electrons with the sample forms the image. Local charging of the sample, along with the intensity of the electron beam and the position of the detector, create apparent lighting and shadowing effects in the electron microscope image. Further, an SEM can resolve features much smaller than the wavelength of light, and has a relatively large depth of focus.
The images can be digitally-enhanced, such as by adding colors and highlighting using Adobe Photoshop, as shown in the gallery of colorized images.
When the desired field of view exceeds the field of view of the electron microscope (typically a few millimeters), several frames can be "stitched" together to effectively create a wide-field electron image. This technique was demonstrated using Microsoft Expression Graphic Designer, in collaboration with Michael Cohen at Microsoft Research and Felice Frankel at MIT. More examples of the stitching technique are in the gallery of stitched images.
Figure 5. Images of "Seed of Life" pattern of carbon nanotube structures grown on a silicon substrate: (left) digitally-stitched raw SEM image; (middle) colorized SEM image; (right) optical photograph. This arrangement of circles having 6-fold symmetry was first found at the Temple of Osiris at Abydos, Egypt.
References
[1] A.J. Hart. Chemical, Mechanical, and Thermal Control of Substrate-Bound Carbon Nanotube Growth. Ph.D. Thesis, Massacusetts Institute of Technology, Cambridge, MA, August 2006 (view extended abstract).
[2] A.J. Hart and A.H. Slocum. Rapid Growth and Flow-Mediated Nucleation of Millimeter-Scale Aligned Carbon Nanotube Structures from a Thin-Film Catalyst (with cover feature), J. Physical Chemistry B 110(16):8250-8257, 2006, http://dx.doi.org/10.1021/jp055498b.
[3] A.J. Hart and A.H. Slocum. Force Output, Control of Film Structure, and Micro-Scale Shape Replication by Carbon Nanotube Growth Under Mechanical Pressure, Nano Letters 6:1254-1260, 2006, http://dx.doi.org/10.1021/nl0524041.
[4] A.J. Hart, H.K. Taylor, and A.H. Slocum. Three-Dimensional Growth of Carbon Nanotubes on Substrates: From nm-to mm-Scales, 4th International Symposium on Nanomanufacturing, Cambridge, MA, 2006 (download manuscript).
[5] R.D. Bennett, A.J. Hart, and R.E. Cohen. Controlling the Morphology of Carbon Nanotube Films by Varying the Areal Density of Catalyst Nanoparticles Using Block Copolymer Micellar Thin Films, Advanced Materials 18:2274-2279, 2006, http://dx.doi.org/10.1002/adma.200600975.
[6] A.J. Hart, L.C. van Laake, and A.H. Slocum. Desktop Growth of Carbon Nanotube Monoliths with In Situ Optical Imaging, Small 3(5):772-777, 2007, http://dx.doi.org/10.1002/smll.200600716.
 
Home
c.2007 - ajohnh@umich.edu