Growing CNTs by chemical vapor deposition

Electron beam evaporation of pure iron is an excellent way to deposit catalyst. If a thin layer is used (0.1 -0.4 nm) then you can grow exclusively single wall CNTs. We use the recipe published by the Rogers group (Nature Nanotech 2007). It is useful to look at the McEuen group's version of this recipe.

A 0.2 nm layer of Fe is invisible when you inspect the chip by naked eye, but you can see the iron in an AFM (image below). The Fe layer will bead up on the surface at high temperature and yield nanotube tube diameters < 2 nm.

Figure: 20×20 micron AFM scan showing CNTs growing from a stripe of evaporated iron (10 microns wide and 0.1 nm “thick”) deposited on ST-cut quartz. Patterning was done by photolithography. Nanotubes can be seen growing above and below the iron stripe. Growth direction is perpendicular to the wafer flat.

We have tried at least three different iron-based catalysts that are suspended in a solvent. Even more variations exist in the literature. The Alumina supported catalyst is tried and true:

Figure: SEM image of CNTs growing from alumina supported catalyst. Scale bar is 200 nm. The advantage of alumina supported catalyst is the high surface area of the alumina. The high surface area leads to more nanotube nucleation sites. One disadvantage is that alumina nanoparticles, together with sticky nanotube grass, stick to the AFM tip during imaging (it is best to avoid imaging directly over the alumina!). A second disadvantage is that the diameter distribution of CNTs tends to be broad.

1. Alumina supported catalyst. The original recipe for this catalyst was published in Nature in 1998 (Kong et al.). Kong used 45 mL of IPA instead of 45 mL of DI water, but IPA is not compatible with photoresist patterning (IPA is only compatible with ebeam resist). The original recipe from Kong et al. also has an additional baking step after solution deposition. Here is the updated recipe developed in the McEuen group. The Dekker group still uses the IPA solvent:
   • 60 mg Fe(NO3)3•9H2O * see the suggestion below
   • 15 mg MoO2(acac)2 * see the suggestion below
   • 45 mg Al2O3 (Alumina)
   • 45 mL DI water
   • Stir (10 min with magnetic stir rod) and then sonicate several hours when first made, should look red ()
   • * Suggestion Our current formula has four times as much Iron Nitrate and Molybdenum Acetate as the preceding formula. This was done to increase our return on the number of nanotubes. We know the new formula does grow tubes, but it does not look like it makes more tubes than the recipe at lower concentrations like we hoped. Another potentially important difference is that we are using Mo(acetate)2. Where Mo is in the 2nd oxidation state. The MoO2(acetate)2 is in the 6th oxidation state. We tried growing tubes using a solution made exclusively of of Mo(acetate)2 and alumina. We did not find any tubes grown using this solution. So it looks like all tube growth (so far) is due to either Fe or an FeMo alloy.
2. Iron nitrate solution
   • Mix 60 mg Fe(NO3)3•9H2O, clean with 45 mL DI or IPA (concentration calculation). When working with a water based catalyst there are some additional considerations.
   • Either drop cast or dip coat the chip.
3. Ferritin solution
   • Always keep bulk supply of Ferritin in the fridge at ~ 4deg C
   • Dilute bulk supply at least 1:100 in DI water (Josh has details)
   • To get a reliable and uniform distribution of Ferritin the solution should be spun onto the chip and the chip should be pre-treated to get a hydrophilic surface.

The CVD furnace uses quartz tubing to contain the gas flow while everything (chip, catalyst, gas) is heated to high temperature. We have spare quartz tubes in the event that the tube becomes contaminated. We have no evidence that quartz tubes become contaminated by a reliable growth recipe, or that alumina support catalyst solution becomes too “old”.

THIS RECIPE WAS RECENTLY UPDATED (09/21/2009)

(Click here for other recipes)

• Check that the gas cylinders are all on and have plenty of gas (the pressure inside a full cylinders is about 1500 psi).
• Open furnace lid. Open growth tube carefully. Place sample within growth tube. Use the metal push rod to slide the sample to the center of the furnace.
• Close growth tube. Establish vacuum seals to the quartz tube. Close furnace lid.
• Set growth parameters in automated growth program.

• Alcohol Growth Parameters

1. Anneal temperature = 800C
2. Anneal time = 15 minutes
3. Anneal gas = Atmosphere
4. Growth temperature = 900C
5. Growth time = 15 minutes
6. Growth gases = 0.45 SLM Hydrogen, 0.30 SLM Methanol, and 0.15 SLM Ethanol

• Methane Growth Parameters (suggested by McEuen group)

1. Anneal temperature = 550C
2. Anneal time = 5 minutes
3. Anneal gas = Atmosphere
4. Growth temperature = 825C
5. Growth time = 60 miniutes
6. Growth gases = 1.9 SLM Methane, 0.3 SLM Hydrogen

• Make sure furnace and power supplies are on.
• If you are doing alcohol growth add ice and water to the beakers that the bubblers are in.
• Press the start button on the automated growth display panel.
• Allow system to cycle.
• At 400C, safe to open furnace lid (speeds up cooling process)
• At 200C, safe to handle growth tube and retrieve sample.

Gases are ordered from Industrial Gas and Welding Supplies, Corvallis:

• Methane: “Chemically Pure Minimum Purity” 99.0% Airgas Part #: ME CP300
• Hydrogen: “Ultra High Purity Minimum Purity” 99.999% Thc < 0.2 PPM Airgas Part #: HY UHP300
• Argon: “Ultra High Purity Minimum Purity” 99.999% Thc < 0.5 PPM Airgas Part #: AR UHP300