Page 53 - Winter2018

AB
platinum wires or gold leaf, although this efficiency is still well below that of most conventional transducers.
Although CNT sheets are considered to be very mechani- cally robust by certain metrics (and can support droplets of liquid 50,000 times the weight of the sheet itself), their ex- treme porosity leaves them vulnerable to damage by even small macroscopic natural events such as a droplet of water falling on the sheet or a moderate gust of air. In an attempt to remedy this problem, other thermophone heaters that have been manufactured and examined include graphene sponges (Fei et al., 2015), CNT sponges (Aliev et al., 2015), carbonized electrospun polymers (Aliev et al., 2016), and carbon fiber (Dzikowicz et al., 2017). These denser but also more manageable materials highlight the tradeoff between mechanical robustness and thermoacoustic efficiency.
Another parameter thermophone designers must take into consideration is device scalability, which asks how quickly and efficiently these materials and devices can be manufac- tured and assembled in a repeatable fashion. Since Shinoda et al. (1999), various other thermophone active elements have been produced using VLSI technology such as multi- layer (Tian et al., 2011a) and single-layer (Suk et al., 2012) graphene sheets, tungsten thin films deposited by atomic layer deposition (Brown et al., 2016), and thin gold (Dutta et al., 2014), silver (Tian et al., 2011b), and aluminum (Nis- kanen et al., 2009) wires.
The group from Tsinghua University who published the first CNT-based thermophone element described, in a patent submitted shortly thereafter, sound produced from such a device when submerged just beneath the surface of water (Jiang et al., 2008). Aliev et al. (2010) drew the conclusion that such an effect is possible underwater because carbon nanotubes are hydrophobic and sustain a thin layer of air that then thermally expands on heating as opposed to rely- ing on the thermal expansion of water, which is quite negli- gible below vaporization. Although sound can be discerned from a pristine CNT thermophone that has been submerged just below the surface of the water, removing the CNT sheet from the water results in physical damage to the element.
An alternative to producing more robust thermophone ele- ments is to shield the fragile element from the external envi- ronment. To protect CNT sheets, Aliev et al. (2010, 2014b) and Mayo (2015) have encapsulated them between various materials ranging from polyimide film and mica sheets to ceramic and metal plates. Introducing the encapsulation media results in a mechanical system with resonances that
E
C
D
Figure 3. A: 1–meter-long freestanding carbon nanotube (CNT) sheet pulled from the edge of a CNT forest by Zhang et al. (2005). B and C: scanning electron microscope images of the interface between the horizontally aligned sheet and the vertically aligned forest, re- spectively. D: layering or stacking of CNT sheets at various angles. E: hydrophobic CNT sheet can support droplets from various aqueous solutions. Images reprinted from the AAAS, with permission.
tinuously drawn from the forest and, if desired, spun into fibers. Individual CNTs can be semiconducting or metallic, depending on their chirality (i.e., the relative orientation of the 2-D lattice to the angle in which the lattice is “wrapped” on itself), but, statistically speaking, a random array of vari- ous chirality CNTs is electrically conducting (Saito et al., 1998). The high porosity of CNT sheets affords them a much larger interface with the surrounding gas and results in more efficient acoustic thermophones than those that utilized thin
Winter 2018 | Acoustics Today | 51