Device fabrication and characterization:
Even though the CVD method works well to produce random networks of single wall carbon nanotubes, the procedure has serious disadvantages: the fabrication process is incompatible with standard CMOS techniques, and there is a lot of catalyst junk on SiO2 surface and nanotubes. Our goal is to fabricate clean nanotube network transistors at room temperature on various substrates (transparent, flexible, silicon) using a liquid phase deposition method.
Fabrication methods:
One method we use for making nanotube networks on various substrates is a spin coating technique, first established by Meitl et. al. This method uses a sonicated solution of nanotubes in Sodium Dodecyl Sulfonate, which can separate bundles and encapsulate them as single nanotubes in solution. This solution is than centrifuged, to remove larger bundles and catalyst material.
Before and after Centrifuging. The dark solution on the right contains large bundles and catalyst material that form a dark pellet at the bottom of the vial on the left.
The centrifuged material is than spin coated onto a substrate, along with methanol. The methanol acts to remove the SDS from the nanotubes, and the result is a film of single tubes to small bundles. This method can be used to coat a variety of substrates, such as silicon dioxide, Mica, and mylar. We have found it helps to get a dense network if you first treat the surface with a solution of APTES, so that nanotubes can stick to the amine groups. Using this technique, nice films on PET have been made. The result is both conductive (dependent on volume of solution used in spin coating), and transparent to optical wavelengths. With some improvements in nanotube chemistry, this may become a replacement for ITO coatings.
Spin coating process
NT Spun on Polyethylene (PET)
Nanotube networks can be coated onto almost any substrate you wish using a simple spraying technique. A solution of nanotubes is pressurized with air and forced out through a small opening in a glass beaker. This creates an ultra fine mist that can be deposited onto most any substrates. Networks on transparent substrates such as glass and Polyethylene have been made, with the sheet resistance and transparency controllable by the volume of solution sprayed. A spray coating method may be useful in the future for industrial applications.
NT sprayed onto glass slide (right) vs. blank slide (left).
Filter-based nanotube network.
Fabrication method: we used a vacuum filtration method, which involves vacuum filtering a dilute suspension of nanotubes in a solvent over a porous alumina filtration membrane. As the solvent falls through the pores, the nanotubes are trapped on the surface of the filter, forming an interconnected network. The density of this network (nanotubes/area) can be controlled to high precision, to make either quite rare, sub-monolayer networks, or dense, thick networks as desired. Our method has the major benefit that the quickness of the vacuum filtering process does not allow for tube flocculation, creating optically homogenous films.
SEM images:
a) V=7 ml (near to percolation) b) V=10 ml (just above percolation) c)V=400 ml (well above percolation)
Device application: The nanotube network on filter is ideal for liquid-gating transistor and bio/chemical sensing as well.
Flexible transparent nanotube network.
Fabrication method: Nanotubes are first vacuum filtered onto an alumina membrane as described above. These nanotube network filters are then covered with an evaporated layer of a polymer called Parylene C. This polymer is an industrial standard as an dielectric layer, and is very transparent as well. After the parylene C is deposited, it can be peeled off, removing the nanotube network along with it. This procedure can provide a room temperature mechanism for obtaining a transparent, conductive, flexible network, where the density of the network can be finely controlled.
Device application:
Transparent and/or flexible transistor, field emission device and solar cell.