Thermal fluctuations, stress relaxation, and actuation in carbon nanotube networks

Do carbon nanotubes behave like polymers? The answer appears to be—some do and some do not, depending on the number of walls. Analyzing the behavior of carbon nanotube networks, as found in sheets of “bucky paper,” provides an intriguing insight into the characteristics of nanotubes, noninvasively deducing the fundamental response of individual tubes from the average characteristics of the collective. We report stress relaxation experiments on single- and multiwalled carbon nanotube networks and also present their reversible photomechanical actuation response. Experimental similarities between multiwalled nanotube networks and a “sticky” granular system are observed, while single-walled tubes display entropic behavior akin to a polymer network. In both cases, photostimulated actuation is orders of magnitude larger than thermal expansion predictions. The analogy between single-walled tubes and entropically driven polymer chains, and between multiwalled tubes and granular networks, suggests a paradigm for theoretical and experimental analysis.

Figure 1

Networks of carbon nanotubes. (a) SEM image of MWNT mat. (b) SWNT mat, under similar magnification, highlighting the tube bundles. (c) Scheme of the experiment, which controls the temperature $(T_1,T_2)$ and strain via a micrometer (M) and measures the exerted force via a dynamometer (D), in ambient relaxation or on irradiation.

Figure 2

(Color online) Stress relaxation of a MWNT mat kept at fixed length, after a step strain of 0.2%. Different curves obtained in the same experiment at 28, 41, 60, and $90°\mathrm{C}$ are color coded. The inset shows the relaxation during the first $5\mathrm{h}$.

Figure 3

(Color online) The analysis of the long relaxation in a MWNT mat carried out at $41°\mathrm{C}$. The data are fitted by the power law $\Delta \sigma \left(\mathrm{MPa}\right)=0.5+0.9t^{-0.2}$ and a logarithmic decay $\Delta \sigma =1.3-0.2\ln \left(t\right)$. The inset plots the same data on a $\ln t$ axis, indicating that logarithmic relaxation (labeled by ◻) is a much closer fit than the power law (labeled by 엯).

Figure 4

(Color online) Stress relaxation of a SWNT mat kept at fixed length, after a step strain of 0.2%. The jump in stress is different at each temperature. However, the inset displays the normalized data $\Delta \sigma ∕\Delta {\sigma }_{\max }$, indicating the universal relaxation mechanism, not altered with temperature.

Figure 5

The variation of Young’s modulus in a SWNT mat with temperature measured at three separate times of relaxation.

Figure 6

(Color online) Photomechanical actuation of a MWNT mat recorded at fixed, prestrained, and equilibrated sample length. The inset shows the initial contractive stress response of the film during the first few seconds when the light source is switched on, illustrating the initial peak.

Figure 7

(Color online) Photomechanical actuation of a SWNT mat, in the same conditions. In contrast to MWNT network, this sample is contracting on illumination. The detailed onset kinetics, highlighted in the inset, matches well the compressed exponential result reported in Ref. 9.