Researchers have developed a way of “wiring up” graphene nanoribbons (GNRs), a category of one-dimensional supplies which might be of curiosity within the scaling of microelectronic gadgets. Utilizing a direct-write scanning tunneling microscopy (STM) primarily based course of, the nanometer-scale metallic contacts had been fabricated on particular person GNRs and will management the digital character of the GNRs. The researchers say that that is the primary demonstration of creating metallic contacts to particular GNRs with certainty and that these contacts induce system performance wanted for transistor perform.
The outcomes of this analysis, led by electrical and pc engineering (ECE) professor Joseph Lyding, together with ECE graduate scholar Pin-Chiao Huang and supplies science and engineering graduate scholar Hongye Solar, had been lately printed within the journal ACS Nano.
“Graphene has been round for some time and it has been considered one thing that would probably be a high-speed digital materials, even perhaps a alternative for silicon,” explains Lyding. “However the issue with graphene itself is that it isn’t a semiconductor.” Graphene is a one-atom-thick layer of carbon atoms and whereas it could be the thinnest recognized materials, additionally it is extremely sturdy. Semiconductor properties may be induced in graphene by making it very small or by fabricating it into particular shapes — like ribbons. For this mission, atomically-precise GNRs had been synthesized by co-author Alexander Sinitskii and his group on the College of Nebraska.
The method of creating a transistor out of the GNRs contains placing them on a silicon substrate, connecting wires and working present by way of the wires to measure the transistor properties. The crew has made the vital step of taking the GNRs, which might be narrower in diameter than a DNA molecule, and wiring them up. They’ve developed a way the place the wires are additionally just some nanometers extensive.
Different researchers have labored on this downside by placing many GNRs on a silicon floor and placing down large electrodes and hoping for the very best. This methodology, nonetheless, introduces loads of uncertainty. Lyding and his college students used a extra exact methodology for wiring up the GNRs. They used a scanning tunneling microscope (an atomic decision imaging software) to scan the floor on the lookout for a GNR to make use of. In STM, a pointy tip is introduced near a floor — on the order of a nanometer — and scanned throughout the floor. There’s a present circulation between the tip and the floor, and when the tip comes throughout atoms on the floor, like driving over a speedbump, that present circulation turns into modulated. This enables for the detection and imaging of the GNRs.
As soon as they discover a GNR, they use the electron beam within the STM to set off metallic deposition from hafnium diboride precursor molecules to create the wires. Co-author Gregory Girolami and his group within the UIUC Chemistry Division synthesized the precursor for this course of, known as STM direct-write. “Our wiring methodology could be very exact. After we see a GNR, we will simply outline a sample that we would like, after which we’ll join it. It isn’t simply blindly throwing electrodes on the floor,” says Huang.
One other benefit of this methodology is that it’s finished in ultra-high vacuum (UHV). This ensures that the fabric stays clear from atmospheric water and different “junk” that degrades system efficiency.
The researchers additionally investigated the digital character of the GNRs and located that it was modified by placing the metallic contacts on. Semiconductor “doping” is the intentional introduction of impurities to vary its digital properties. Solar explains, “One option to dope GNRs is to make use of totally different chemical reactions to vary the GNR properties. However that course of is difficult. The best way we do it’s by depositing metallic. And we will truly select the type of metallic that we wish to placed on the GNRs which might additionally tune the GNR traits. That is one option to basically dope our GNRs, with out truly utilizing dopants.”
Lyding says, “The subsequent step, which we’re engaged on now, is to make an actual transistor and really measure the transistor traits. However we all know that we will do that pristine course of, utilizing ultra-high vacuum, of creating the electrodes which might be completely obligatory for system perform.”