Growth of vertically aligned Carbon Nanotubes - NanoLabStaff/nanolab GitHub Wiki
by Jon Oddvar Kolnes ([email protected])
The substrate used was a 2 inch, p-type silicon wafer of <100> orientation, with a thickness of about 350µm. Samples are scribed using the Dynatex DX-III scriber unit.
The sample is cleaned by immersion in acetone for 1 minute, isopropanol for 1 min, followed by a rinse in DI water and drying by nitrogen-gun. Any residual liquid is removed by baking on a 180ºC hot plate for 90 sec.
An oxygen plasma clean is performed using the Diener Electronics Femto Plasma Cleaner, with a gas flow of 100 sccm O2 and a 50W power level for 2 minutes (50%/50%).
To grow the carbon nanotubes, a barrier layer is needed to avoid diffusion between substrate and catalyst. Most oxides can be used, in this particular method description aluminium oxide was used. It was deposited using the AJA Sputter.
A deposition of aluminium in an atmosphere of 60 sccm Ar, and 6 sccm O2 results in a layer of aluminium oxide. A layer of 100 nm was deposited. The thickness may be confirmed by using the Filmetrics F20 Reflectometer.
A 3nm iron catalyst layer was then deposited. Thicknesses between 1nm and 10nm will give carbon nanotubes. Other catalyst metals such as Ni and Co may also be used.
The carbon nanotubes were grown using the Oxford Instruments PlasmaLab System 100-PECVD. The recipe used was the 'CNT N2O Heat Pretreatment'. A recipe developed by Espen Rogstad based on the work of previous master's students that have been growing CNTs using this PECVD.
Resting on the 4 inch silicon carrier wafer, the sample was inserted into the PECVD chamber at 650°C. A 2 min purge in 1500 sccm N2, followed by a 3 min pump, readied the chamber before pumping to a base pressure of 5×10^-6 mTorr. A 30 min thermal pretreatment period with 50 sccm N2O prepared the chamber before introducing the hydrocarbon gas. For 120min, a gas flow of 50 sccm CH4 was used to grow the carbon nanotubes, with a 100W plasma power. A new series of N2-purging and pumping was then repeated three times while the temperature returned to 300°C, ending with a final pumping to base step. The pressure was kept at 1000mTorr during pretreat- ment and CNT growth, and at 1500mTorr during the nitrogen purge.
A colour change is observed during processing as observed in the image below. From the clear silicon surface, the surface turns blue after deposition of the 100 nm aluminium oxide layer, and then black after growth of carbon nanotubes.
For cross-sectional imaging in the Hitachi S-5500 S(T)EM, the samples should be scribed after depositing the catalyst layer, before growth of carbon nanotubes. This is to avoid any damage of edge-tubes, allowing for better cross-sectional images.
For reasons unknown, the PECVD would often not grow carbon nanotubes. Sometimes, no carbon was observed on the sample after processing, other times, the distribution was not uniform on the processed surface (see image below).
Carbon nanotubes were grown by this method, but these structures varied greatly from each PECVD process. Samples with the same barrier and catalyst layer thickness, would when processed with the PECVD (same settings) return completely different. The height of the tubes ranged from 500 nm to 5 µm. Most of the time, an accumulated carbon layer could be observed on top. This is not desired, as it covers the whole carbon nanotube structure and removes the whole purpose of the surface. It is possible to grow a suffcient surface structure, but the PECVD recipe requires a proper overhaul.
All images have been taken at the NTNU Nanolab by Jon Oddvar Kolnes. Most of them can be found in my Master's Thesis "Fabrication and characterisation of a vertically aligned Forest of Carbon Nanotubes towards a superhydrophobic Surface".