Although the intrinsic variability in nanoelectronic devices has been a major obstacle and has prevented mass production, this natural stochasticity can be an asset in hardware security applications. Herein, we demonstrate a true random number generator (TRNG) based on stochastic carrier trapping/detrapping processes in randomly distributed carbon nanotube networks. The bitstreams collected from the TRNG passed all the National Institute of Standards and Technology randomness tests without post-processing. The random bit generated in this study is sufficient for encryption applications, particularly those related to the Internet of Things and edge computing, which require significantly lower power consumption.INDEX TERMS carbon nanotube network, random number generator, stochastic carrier trapping

As a new type of one-dimensional nanomaterial, carbon nanotubes (CNTs) have been used as a chemical sensing material due to their excellent electrical, mechanical and chemical properties. Several recent studies have attempted to obtain an electrochemical sensor based on CNTs, and CNT-based humidity sensors have potential applications in industrial, agricultural and medical fields and in smart wearable electronic devices. Although various CNT-based humidity sensors have been reported, in this work, we investigated the humidity effects in depth and the underlying mechanisms according to CNT type, i.e., semiconducting (s-CNT) and metallic (m-CNT) CNTs, which is determined by the chiral vector during CNT growth. For this purpose, we fabricated CNT-based humidity sensors with highly purified, solutionprocessed 99% single-walled s-CNT and m-CNT network films bridged by palladium (Pd) electrodes. The fabricated sensors exhibited completely different humidity responses, as denoted by the film resistance as a function of the relative humidity (RH), according to the CNT type. In the s-CNT-based sensor, the resistance tended to decrease as the RH increased, while the m-CNT-based sensor showed the opposite tendency. Based on these results, a humidity sensing mechanism according to the CNT type was proposed in this work. We believe that our findings can serve as design guidelines for CNT-based humidity sensors.

Highly purified, preseparated semiconducting carbon nanotubes (CNTs) hold great potential for high-performance CNT network transistors due to their high electrical conductivity, high mechanical strength, and room-temperature processing compatibility. In this paper, we report our recent progress on CNT network transistors integrated on an 8-inch wafer. We observe that the key device performance parameters of CNT network transistors at various locations on an 8-inch wafer are highly uniform and that the device yield is impressive. Therefore, this work validates a promising path toward mass production and will make a significant contribution to the future field of wafer-scale CNT electronics.

Resistive crossbar arrays can carry out energy-efficient vector-matrix multiplication, which is a crucial operation in most machine learning applications. However, practical computing tasks that require high precision remain challenging to implement in such arrays because of intrinsic device variability. Herein, we experimentally demonstrate a precision-extension technique whereby high precision can be attained through the combined operation of multiple devices, each of which stores a portion of the required bit width. Additionally, designed analog-to-digital converters are used to remove the unpredictable effects from noise sources. An 8×15 carbon nanotube transistor array can perform multiplication operation, where operands have up to 16 valid bits, without any error, making in-memory computing approaches attractive for high-throughput energy-efficient machine learning accelerators.

Carbon nanotubes (CNTs) are one-dimensional materials that have been proposed to replace silicon semiconductors and have been actively studied due to their high carrier mobility, high current density, and high mechanical flexibility. Specifically, highly purified, pre-separated, and solution-processed semiconducting CNTs are suitable for mass production. These CNTs have advantages, such as room-temperature processing compatibility, while enabling a fast and straightforward manufacturing process. In this paper, CNT network transistors were fabricated on a total of five 8-inch wafers by reusing a highly purified and pre-separated 99% semiconductor-enriched CNT solution. The results confirmed that the density of semiconducting CNTs deposited on the five selected wafers was notably uniform, even though the CNT solution was reused up to four times after the initial CNT deposition. Moreover, there was no significant degradation in the key CNT network transistor metrics. Therefore, we believe that our findings regarding this CNT reuse method may provide additional guidance in the field of wafer-scale CNT electronics and may contribute strongly to the development of practical device applications at an ultralow cost.