US researchers say that RF heating is ten times faster than two-point-probe measurements

Carbon nanotubes (CNTs) were first synthesised almost thirty years ago, in 1991. In the decade that followed, researchers established many of their properties, and took the first tentative steps towards integrating them into working electronic devices. The development of solution-based printing techniques for CNTs in more recent years marked a major change of pace, allowing CNT structures to be integrated into everything from sensors and solar cells, to thin-film transistors and antennas. But while it’s now possible to fabricate CNT-based devices quickly and at scale, quality assurance tests for devices, such as two- or four-point-probe resistance measurements, remain slow and inefficient. Researchers at Texas A&M University have suggested an alternative screening approach for these devices – one based on the thermal properties of CNTs.

In previous work, the team had shown that CNTs heat up when placed in low radio frequency electric fields (1 – 200 MHz). They also showed that the heating rate depended on the exact frequency used, and on the electrical properties of the CNTs. So, they reasoned that this RF heating phenomenon could be adopted as a high-throughput screening tool for nanotube devices. They reported their findings in an upcoming issue of Carbon [DOI: 10.1016/j.carbon.2019.06.039].

The experimental setup was relatively simple – the heater was made from copper tape laid on a Teflon base. One end of each strip of tape was connected to the RF source, generating a fringing-field which extended vertically out of the plane of the base. This was mounted on a motorised linear stage that could move commercially produced sheets of CNT circuits through the RF field, while being imaged by a top-mounted infrared camera.

The RF heating caused a temperature change of about 4 °C in each circuit, so was too low to cause any degradation to the device. However, it was sufficient to identify multiple faults in some circuits, including gaps, and areas of non-uniform CNT density. The same sheets were also tested using a traditional two-point-probe. In most cases, both the RF heating and electrical resistance tests could identify faulty circuits. However, the traditional approach “passed” a small number of faulty circuits, and it was significantly slower. Scanning one sheet of 377 circuits via the RF heating method took approximately 30 s, compared to close to 10mins for the two-point-probe tests.

They applied their same experimental setup to devices with silver electrodes, and other circuits based on carbon black. The results were similarly positive, leading the authors to suggest that it could be applied to any type of RF-responsive material, including graphene. They say that their method “…provides a tenfold increase of speed, allowing for real-time monitoring of circuits as they are produced…. (and it) is more reliable than the traditional metric in screening-out faulty circuits.”


Victoria K. Hicks, Muhammad Anas, Erin B. Porter, Micah J. Green. “High-throughput screening of printed carbon nanotube circuits using radio frequency heating”, Carbon 152 (2019) 444-450. DOI: 10.1016/j.carbon.2019.06.039