“Nanochannels are essential components in nanofluidic systems,” Choonsup Lee told nanotechweb.org. “[Nanochannel manufacturing methods] should be cost effective, able to precisely control channel dimensions and CMOS compatible for ultimate integration with microelectronics. One of the easiest ways to make a nanochannel is to use e-beam lithography, but it is extremely expensive so I decided that we needed a cheap and simple nanochannel fabrication technique.”
To form the nanochannels, Lee and colleagues added a 100 nm layer of amorphous silicon to a silicon substrate. The thickness of this layer determined the depth of the nanochannel. The researchers used photolithography to assign the length and shape of the nanochannel, etching away portions of the amorphous silicon with reactive ion etching and then oxidizing the sample at 1000°C. The thickness of the resulting oxide layer - in this case about 50 nm - controlled the nanochannel width. In practice, the scientists found that the minimum width for the nanochannels was 5 nm.
The next step was to deposit a 500 nm thick layer of amorphous silicon and then re-expose the oxide layer by chemical-mechanical polishing. Finally, the scientists etched away the oxide layer, producing a nanochannel. They sealed the nanochannel with evaporated gold or silicon oxide.
“These are the longest set of parallel nanochannels - centimetres long and less than 50 nm in diameter - ever made with optical access,” said Lee. “It remains to be seen whether fluid can flow through them.”
Although the nanochannel walls were very smooth, amorphous silicon is hydrophobic, which could cause a problem for fluid flow. With that in mind, in some cases before sealing the channel, the scientists oxidized the channel walls to make their surfaces hydrophilic. This had the advantage of enabling the production of nanochannels narrower than 5 nm. The team also found that sealing the channel with silicon oxide by plasma-enhanced chemical vapour deposition lined the walls with a thin layer of hydrophilic material.
According to Lee, the scientists are planning to use nanochannels to separate molecules based on their physical dimensions. “One of the key advantages [of this technique] is that a very large number of different dimension fluidic channels can be made into a single wafer at low cost,” he said. “Thus, it can separate a mixture of many different-sized molecules based on size.” Lee says he is also trying to flow quantum dots through nanochannels to figure out the flow mechanism inside the channel.
The researchers reported their work in Nano Letters.