By Monica Heger
Close on the heels of a group from The Netherlands' Kavli Institute that became the first to demonstrate DNA translocation through a graphene nanopore last month, researchers from the University of Pennsylvania have done the same, only using a slightly different method to construct their graphene nanopore.
Detailed in Nano Letters last month, the new study highlights both the advantages and disadvantages of the two groups' methods for building graphene nanopores. The methods yielded slightly different but comparable results, said Henk Postma, assistant professor of physics at California State University, Northridge, who is working on building a graphene nanogap sequencing device (IS 3/16/2010).
"It's good work and shows that people are steadily working toward using graphene," Postma told In Sequence.
The Penn team was able to demonstrate the translocation of 15 kilobase pair, a 3 kilobase pair, and 400 base pair fragments of double-stranded DNA through multi-layer graphene nanopores between 5 and 10 nanometers in diameter.
Unlike the Dutch team, which used an exfoliation approach to construct its nanopore (IS 7/20/2010) — scraping a single layer of graphene off graphite — the Penn team used a method known as chemical vapor deposition to grow the graphene on a copper substrate. The advantage of the CVD-based method is that it will be easier to scale up, but the disadvantage is that it doesn't produce as pure a graphene layer as exfoliation does.
The Penn researchers grew the graphene on copper, and then dissolved the copper layer to create a freestanding graphene layer. They then created a silicon-nitride membrane with a 1.5-micrometer hole to use as support for the graphene, placed the graphene over the hole, and then drilled a nanopore through the graphene. They tested nanopores ranging in size from 5 to 10 nanometers in diameter.
The team also added a layer of titanium dioxide to the graphene membrane, which helped improve the signal-to-noise ratio by making the hydrophobic graphene more "wettable," said Chris Merchant, lead author of the study and a postdoc in Marija Drndic's physics lab at the University of Pennsylvania.
Titanium dioxide could also theoretically be used to make smaller nanopores, Merchant said. Other groups have demonstrated that coating silicon nitride with titanium dioxide after drilling the pore helps to shrink the pore. But, in this case the team applied the titanium dioxide before drilling the pore, so it did not have that effect.
Another problem the team had was that its graphene membrane appeared to have pinholes in it, allowing the ionic current to flow not only through the nanopore, but also through the pinholes, which would hinder accurate base reading. The pinholes were likely created in the chemical vapor deposition process, said Merchant.
"When you grow graphene, it doesn't grow in a single crystal way, so you get grains at different thicknesses," said Merchant. "The different grains and thicknesses are affected differently, and we get the creation of pinholes." However, Merchant said that the technology to construct graphene wafers is quickly improving, so he thought the problem could be solved.
Also due to the chemical vapor deposition process, the graphene layer was actually several layers of graphene, so was not as thin as the membrane created by the Dutch group, which was only one single layer of graphene. Improvements to the process should allow them to eventually create a single graphene layer.
While the team successfully translocated DNA through the graphene nanopore, significant hurdles remain. Perhaps the most difficult challenge will be single-base resolution. The problem is "far from trivial," Cal-State's Postma said. "Graphene nanopores are interesting, but I don't think they will take you all the way" to nanopore sequencing.
He said the main challenge will be in creating a shallow enough nanopore. Graphene is a good choice of membrane because a single layer of graphene is only 0.3 nanometers thick — smaller than the distance between two bases. But, Postma said, the addition of the potassium chloride ionic solution that the nanopores are immersed in might make it impossible to create a membrane that is truly only one atom thick. "There is a layer of ions immobilized on the graphene sheet, and because they are not moving, they make the sheet thicker," he said. "I don't know if that can be solved. It's somewhat of an open question."
The researchers, while acknowledging that will be a major issue, are continuing to press ahead. "The effective thickness of the pore is really important, and getting down to a single base will be challenging," said Meni Wanunu, a postdoc at the Penn lab and co-author of the Nano Letters paper. Already, the thickness of nanopores has been reduced from 20 nanometers in silicon nitride membranes to just a few nanometers in the recent graphene nanopores, he added. "We can't say whether we'll be able to do it. But we think that we're moving in the right direction."