Are carbon nanotubes stable enough for use in sensors?

Significance 

A plethora of literature exists that offer plausible information regarding the exceptional performance of sensors based on carbon nanotubes (CNT) transistors, with promises of transformative impact. As such, CNT-based biosensors have received immense attention over the decades, following which numerous demonstrations of sensors that make use of nanotubes as the semiconducting channel in a carbon nanotube field-effect transistor (CNTFET) have been exhibited.

Regardless of such strong research activity, fundamental understanding of how CNTFETs respond when operated under bias stress conditions relevant to many sensor applications is still missing. In fact, little has been done to explore bias stress for CNTFETs with channels of one or a few parallel CNTs, with existing reports primarily focusing on gate bias stress, rather than the combination of gate and drain bias stress that is required for most device applications. This lack of information regarding the effects of bias stress on CNTFETs over long time scales is an obstacle to the field.

In a recent publication, Duke University researchers: Steven Noyce (PhD candidate), James Doherty (PhD Student), Dr. Zhihui Cheng, and Prof. Aaron Franklin in collaboration with Dr. Hui Han and Dr. Shane Bowen at Illumina, Inc. developed a measurement platform that allowed for robust, long-term testing of numerous CNTFETs concurrently in a fully automated manner. In addition, they also established bias ranges under which CNT transistors could operate continuously for months or more without degradation. Their work is currently published in the research journal, Nano Letters.

In brief, the team employed a custom characterization system from which they were able to determine the impacts of defect formation and charge traps on the stability of CNT-based sensors under extended bias. In addition to breakdown (which is well-known), they identified three additional operational modes: full stability, slow decay, and fast decay. These four distinct operational modes were based on applied drain-source bias, wherein devices held at sufficiently low voltage exhibited stability for months of continuous operation. Moreover, they established that as the applied bias increased beyond the full stability range, the CNTs experienced irreversible slow and then fast decay until eventually breaking down.

The researchers also identified a current drift behavior that reduced dynamic range of a CNT-based sensor by over four orders of magnitude but could be avoidable with appropriate sensing modalities. Hence, the importance of operating voltages and related device configuration is high for properly functioning CNTFETs being applied to sensing applications. The structure used for the study was chosen to be similar to a majority of previously studied CNT-based sensors so that the results achieved could have close applicability to the field as a whole.

“Carbon nanotube-based sensors have been studied for decades, but mostly with a focus on unique ways to integrate or apply them for a given application,” said Prof. Franklin whose lab led the research. “This study digs deeper into the operation of the nanotubes themselves when they are held under prolonged electrical stress, which is often required for sensing applications. Our findings show great promise for using the nanotubes, but with certain constraints that must be factored in when determining how to bias them in a sensor.”

In summary, an electrical characterization platform that allows for the long-term interrogation of CNTFETs under various bias conditions was developed. Ultimately, the results presented describe what behaviors are to be expected when CNTs are stressed by continuous and variable biases over long periods of time and demark conditions under which the CNTs are electrically stable, providing valuable insight for the further advancement of the thousands of diverse CNTFET-based sensors that have been proposed. Further, the electrical characterization platform developed in the work is more broadly applicable to other nanoelectronic devices that require long-term study under various bias conditions.

Are carbon nanotubes stable enough for use in sensors? - Advances in Engineering

About the author

Steven Noyce is a PhD candidate in the Electrical and Computer Engineering Department at Duke University. Steven’s passion for discovery triggered the beginning of his research career synthesizing carbon nanotubes in a high school chemistry lab. This work facilitated connections with the university lab of Robert Davis, where Steven did research while earning his B.S. degree in physics, minoring in computer science. Projects in this lab included developing resonant chemical sensors with extremely high surface areas, developing self-assembled circuits using carbon nanotubes and DNA origami, and building high aspect ratio microelectromechanical systems using metal electroplating. During an internship at Cornell University Steven worked on developing a solid state nanopore DNA sequencer, with the result being spun out into a company: Esper Biosciences.

In his graduate work, Steven has focused on building sensors from nanomaterial transistors and determining how these nanomaterials respond to local charge perturbations, changes in environment, etc. This work has spanned many materials including carbon nanotubes, 2D materials such as molybdenum disulfide, and silicon nanowires, while fabrication strategies involved have ranged from traditional clean room approaches to electronics printing. Each of these studies has produced sensor optimization insights that allow for large improvements in nanomaterial-based biosensors, ultimately contributing to low-cost and high-precision medical diagnoses.

On the side, Steven is a founder and the Director of Engineering for CattleGuard Inc., a startup focused on using Internet of Things technology to improve agricultural efficiency.

Outside of work, Steven enjoys rock climbing, spending time outdoors with family, playing the cello, tinkering with circuits, and toying with machine learning. Additionally, Steven gets a thrill out of seeking to ignite a spark of discovery and exploration in his wife and two children.

About the author

Dr. Aaron Franklin is the J. L. & E. M. Vincent Associate Professor of Electrical & Computer Engineering at Duke University. He received his Ph.D. in Electrical Engineering from Purdue University in 2008 and then spent six years on the research staff at the IBM T. J. Watson Research Center in Yorktown Heights, NY. He is most widely known for his work on low-dimensional nanoelectronics with specific emphasis on carbon nanotube (CNT) transistors, including device scaling, transport studies, and advanced integration approaches. Dr. Franklin is a prolific scientific contributor, with more than 70 publications, 50 issued patents, and numerous contributions to international conferences. He joined the Duke faculty in 2014.

Research in Dr. Franklin’s lab is focused on improving the performance and functionality of nanomaterial-enabled electronic devices. This includes high-performance devices from low-dimensional materials such as 2D crystals, carbon nanotubes, and nanowires. Also included is the low-cost realm of printed electronics, which benefits from the incorporation of nanomaterials to enhance electrical transport over large printed features, among other application advantages. The primary drive of the group’s research is to improve performance for all electronic devices, including those with more custom form factors (flexibility, transparency, biocompatibility, etc.). There are a growing variety of new electronics applications that nanomaterials are uniquely capable of enabling and the Franklin group works to make such applications possible.

Reference

Steven G. Noyce, James L. Doherty, Zhihui Cheng, Hui Han, Shane Bowen, Aaron D. Franklin. Electronic Stability of Carbon Nanotube Transistors Under Long Term Bias Stress. Nano Letters 2019, volume 19, page 1460−1466.

Go To Nano Letters

Check Also

Featured Video Play Icon

Innovation of new Nussbaum functions: adaptive control with unknown direction and time-varying impact