Tsinghua University, Beijing, China
Researchers at Tsinghua University have developed an intelligent artificial throat based on laser-induced graphene. The intelligent device both detects and generates sound in a single unit. The biocompatible artificial throat, which attaches to the larynx, is designed to produce recognizable and controllable sounds. According to the researchers, led by Prof. Tian-Ling Ren in the university's Institute of Microelectronics, the device is a new application of graphene in the field of wearable electronics and is expected to have a significant impact in biology, speech recognition, and other fields.
Acoustic devices include both sound sources and sound detectors; however, these traditional sound sources and detectors, which work in the audible domain (20 Hz-20 kHz), are usually single discrete devices and thus are unable to both generate and detect sound. In addition, traditional acoustic devices are not typically constructed of flexible materials, and so they are not suitable for wearable applications.
How It Works
The integrated acoustic device generates sound based on the thermoacoustic effect of graphene and detects sound based on its piezoresistive effect. The team used a low-cost laser with direct-writing technology to convert polyimide into large-scale patterned porous graphene. The porous graphene has high thermal conductivity and low heat capacity, so it can emit sound by thermoacoustic effect.
In addition, the porous structure of the graphene is extremely sensitive to pressure and thus recognizes weak vibrations on the throat, resulting in sound detection by piezoresistive effect. This piezoresistive effect enables the device to detect humming sounds that can be converted into precise sounds with controllable frequency and intensity. The research was published in an article titled, “An Intelligent Artificial Throat with Sound-Sensing Ability Based on Laser Induced Graphene,” in the journal Nature Communications.
According to the article, when working as a sound source, the laser-induced graphene artificial throat can generate wideband sound with frequency from 100 Hz to 40 kHz, while a thinner laser-induced graphene produces a higher sound pressure level. When working as a sound detector, the artificial throat exhibits unique responses to different sounds and throat vibration modes. The laser-induced graphene can recognize a cough, a hum, and a scream with different tones and volumes. It can also recognize words and sentences. Different volumes and frequencies are converted into controllable, predesigned sounds. Because the laser-induced graphene can be acquired by using a portable, low-power laser platform, risk is reduced.
The group performed tests with six audio types, including a firecracker, a cow, a piano, a helicopter, a bird, and a drum. A 25-m-thick polyimide was chosen to generate the laser-induced graphene because of its resistance change compared with polyimide with a thickness of 75 and 180 m. At 25 m thick, the sensitivity is high enough to detect sound pressure produced by a loudspeaker. The 25-m polyimide was fixed by two free-standing clamps, and the loudspeaker was placed 3 cm away from the artificial throat.
Although the sampling frequency of the artificial throat was 100 Hz, which is far lower than the frequency of sound, the group noted that the responses of the transducer were synchronous to the original audio signals, and the characteristic peaks were retained and reflected with high fidelity. With volume increases, the vibration strengthened, causing a more distinct resistance change. The device showed excellent durability under high strain.
Lu-Qi Tao and He Tian, a PhD student in Ren's group, are the co-first authors of the article. Ren and Prof. Yi Yang are the corresponding authors. This research was funded by the National Natural Science Foundation of China and the Ministry of Science and Technology of China. Ren's research focuses on practical application of graphene devices, particularly on micro-nano electronic devices and flexible graphene sound sources.