Spinning and Processing Continuous Yarns from 4‐Inch Wafer Scale Super‐Aligned Carbon Nanotube Arrays

Abstract

Carbon nanotubes (CNTs) are an extremely interesting type of material due to their unique 1D structure, and their excellent mechanical, thermal, and electrical properties. Many promising applications have been demonstrated. To exploit their excellent physical properties at a macroscopic level, it is desirable to create CNTs with macroscopic length. However, it has been very challenging to grow arbitrarily long CNTs. An alternative approach is to create long nanotube structures with many of them aligned into continuous yarns or ropes. In 2002, a breakthrough was made by our group to fabricate pure CNT yarns by directly drawing CNTs from super-aligned CNT arrays. Following that, Zhang et al. developed a draw– twisting spin method and demonstrated more interesting applications of the raw yarns. Since CNTs in the yarn are nearly parallel aligned, the CNT yarn is intrinsically an anisotropic material and has a special axis along the drawing direction, which demonstrates many fascinating properties and applications such as filaments for light bulbs, polarizers working in the UV region, thermal-field emitters, polarized-light emitters, transparent conducting membranes, etc. As these yarns are macroscopic objects, there is no doubt that they will be sought after for more and more applications as time goes on. However, to achieve real applications in the industry, some key issues have to be solved in advance. Currently many groups have achieved the growth of CNT arrays, but of these only two groups have reported spinning yarns from their arrays. The first question is why our CNT arrays can give rise to CNT yarn while others cannot? What is the critical factor that determines the ability for yarn formation? In our first paper and the recent paper of Zhang et al., the reported syntheses were carried out at atmospheric pressure (AP) in a tube furnace with diameters of 1–2 in. Can this synthesis be expanded to a larger scale at low pressure (LP) for commercial chemical vapor deposition (CVD) systems? The third problem is that, even though many applications have been demonstrated, the directly drawn-out yarns are very sticky due to their clean surfaces and extremely high surfaceto-volume ratio. It can easily to stick to the surfaces of other objects and can never be separated again, which greatly inhibits the real application of CNT yarns. Here, we show how these crucial problems were tackled by our group in the past two years. CNT arrays (or termed “CNT forests”) represent high-quality and highly ordered CNT structures, in which CNTs are nearly parallel aligned and perpendicular to the substrate. To date, a lot of groups have successfully synthesized CNT arrays since the initial papers were published. However, only two groups have reported that CNT yarns can be drawn from their arrays, which were defined as super-aligned arrays by us. We have tried to pull yarns from CNT arrays derived by a previously published method, which is denoted as a normal array, no yarns could be drawn out even though the height of the CNT array was above 1 mm. Now the question is what the difference is between normal and super-aligned arrays? Zhang et al. claimed that the formation of yarn was due to the disordered region at the top and bottom of the CNT array, which entangled together forming a loop. However we cannot agree with them based on the following points. First, in the normal array, the top and bottom part are more disordered and entangled than in the super-aligned array. Second, in our super-aligned array, the bottom part is highly ordered (see supporting information, Fig. S1) and without entanglement. Therefore, the disordered entanglement at the top may help the formation of yarn, but it cannot be the key factor for yarn formation. To elucidate the difference between normal and superaligned arrays, we performed comparative studies. Figure 1A is the transmission electron microscopy (TEM) image of C O M M U N IC A IO N S