Proof of Concepts About Pulling Out Carbon Nanotube Sheets and Yarns—A Molecular Dynamics Study

Although growing carbon nanotube (CNT) forests is well-known today, growing so-called <i>drawable</i> CNT forests, i.e., vertically aligned CNT forests that allow for direct pulling out CNT fibers, sheets or yarns, requires certain conditions. There are three main theoretical hypotheses for the mechanism of conversion of vertically oriented bundles of CNTs into longitudinally oriented CNT fibers during pulling out process. The simplest model [1] considers that the vertically aligned CNT bundles interact only by van der Waals forces. By arguing that van der Waals alone is not able to explain why the CNT bundle being pulled out does not simply detach from the rest of the CNT forest, a second model [2] proposes that something else connects the main vertically aligned CNT bundles in drawable forests. It proposes that individual CNTs or small bundles of CNTs connect the main vertically aligned large CNT bundles, also by van der Waals forces. These connections, during the pulling out of one CNT large bundle, move and get accumulated at the opposite extremity of the forest. This accumulation of connections would, then, transfer the pull-out force on the first CNT bundle to the next, thus keeping the process running continuously. The third model [3] also predicts the existence of connections that get around and entwine one or more large CNT bundles. During the pulling out of the first CNT bundle, the connection strengthens the contact with the neighboring vertically aligned CNT bundle and pulls it out. As the three models might hold true for different structures of CNT forests, using tools of classical molecular dynamics (MD) simulations, we decided to investigate the validity of their main features: van der Waals forces, movement of interconnections, transfer of the pulling-out force from the first to the second large CNT bundle and pulling out of the next bundle. Although the size of real CNT forests prohibits their full atomistic simulation, we designed local structural models that are, at the same time, large enough to test and prove the concept of each feature from the pulling out models, and small enough to be atomistically simulated. We show that the concepts behind the pulling out mechanisms of the last two models hold true, so they are really important in the role of converting vertically aligned CNT forests to longitudinal CNT fibers, sheets and yarns. This work was supported by São Paulo Research Foundation (FAPESP), grant #2020/02044-9.<br/>[1] X. Zhang, <i>et al</i>. <i>Adv. Mater</i>. <b>18</b>, 1505 (2006).<br/>[2] A. A. Kuznetsov, <i>et al</i>. <i>ACS Nano</i> <b>5</b>, 985 (2011).<br/>[3] C. Zhu, <i>et al</i>. <i>Carbon</i> <b>49</b>, 4996 (2011).