SWCNT networks

Figure 26.14 FETs based on SWCNT networks (a) Atomic Force Microscope (AFM) amplitude image of a network of tubes grown by chemical vapour deposition. (b) Gate dependence of conductance of such a network before and after selective ECM. Fjj is 10 mV before and 100 mV after ECM. (c) Variation of conductance of the same device after spotting two different solid polymer electrolytes and using an electrode in contact with the SPE as the gate (F s = 100 mV). The composition of the SPE is identical to that in Figure 26.9. Fpeg is the voltage applied to the polymer electrolyte-gate.

Following these CNT electronic gas sensor studies, many other methods have been explored focusing on the reduction of fabrication cost. Snow et al. demonstrated that a low-density random network of SWCNTs can be fabricated into p-type thin-fllm transistors [Figure 14.7(c)] with a fleld-effect mobility of about 10 cm / Vs and an on-to-off ratio of about 10 [65]. They demonstrated that such thin-fllm transistors can detect dimethyl methylphosphonate (DMMP), a simulant for the nerve agent sarin, at sub-ppb levels [45]. SWCNT network transistors have also been transferred to polymer substrates to form flexible electronic gas sensors [66]. Other resistive sensors based on random SWCNT network [47] or MWCNT films [41] have also been reported. Besides the cost, CNT network and thin-film sensors increase the statistical reliability by averaging out the response at many adsorption sites. This is particularly important when gas concentration is extremely small. 

I-V dependences of SWCNT networks have been presented in some papers... 

In summary, we have shown that the experimental results on the o(7) of SWCNT networks, can be explained by the model based on PhAT initiated by electric field. An advantage of this model over the often used VRH model is the possibility to describe the behavior of I-V data measured at both high and low T with the same set of parameters characterizing the material. On the basis of this model, the phenomenon of the crossover from non-metallic to metallic behavior of the conductivity is explained. The decrease of conductivity at T > Tc in the framework of this model is a result of the temperature dependent carrier tunneling process attended by the phonon emission. 

In this section, the performances of electrical gas sensors based on SWCNTs are reviewed. Three types of gas sensors are introduced field-effect transistor based on individual SWCNT and chemiresistor or field-effect transistor based on SWCNT network. 

Kong et al (2000) have compared the sensitivity to NO2 and NH3 of (a) a CNTFET based on individual semiconducting carbon nanotubes with the sensitivity of (b) a CNTFET-based SWCNT network. The SWCNTs which composed this second device were synthesised by laser ablation. The resistance of device (a) increases and the resistance of device (b) decreases when exposed, respectively, to NH3 and N02.The responses of a CNTFET based on an SWCNT network to 1% of NH3 and 200 ppm of NO2 were approximately 50%. Compared with a CNTFET based on individual SWCNTs, the sensitivity is dramatically reduced. In the case of SWCNT networks, the response of semiconducting SWCNTs is averaged by the small response of metallic carbon nanotubes. Furthermore, due to the van der Waals... 

An integrated chemiresistor based on an SWCNT network was reported by NASA Ames Research Centre (Li et al, 2005). The device was composed... 

Typical response to (a) NH3 and (b) NOj of a chemiresistor based on a SWCNT network (adapted from Battle etal. (2011a)).The dashed arrows indicate the beginning of cooling and the introduction of carrier gas. 

Due to the complexity of the transport properties of an SWCNT network, the physical mechanisms at stake during exposure to gas constitute an important source of debate. The percolation theory is invoked to account for the influence of SWCNT density on the percolation threshold of metallic SWCNTs and on the transport properties of SWCNT networks (Topinka et al, 2009). As a result, the SWCNT network exhibits a metallic or a semiconducting character. 

FIGURE 9.19 (a) Schematic of the SECCM probe and (b) an AFM image of the 2D SWCNT network. (Giiell, A.G., Ebejer, N., Snowden, M.E. et al., Quantitative nanoscale visualization of heterogeneous electron transfer rates in 2D carbon nanotube networks, Proc. Natl. Acad. Sci. USA, 109, 11487-11492. Copyright 2012, National Academy of Sciences, U.S.A.)... 

---