angles were obtained by increasing the twist and decreasing the
draw; conversely, low twist angles were obtained by decreasing the
twist and increasing the draw. Variations in diameter were obtained
by varying the width of the substrates and arrays of CNTs
where the thread was drawn from (Fig. 4). The drawand twist rates
were held constant resulting in similar twist angles in all threads
with varying diameter. The diameter used in calculations was the
average diameter over 20 measurements along the thread imaged
with a Keyence digital microscope.
Since the CNTs were assembled without any foreign solvent
interaction, the resulting thread was fairly porous with many internal
voids, making it difficult to handle and detrimentally
affecting the strength of the thread. To alleviate these problems, the
thread was densified dipping into the organic solvent N-methyl-2-
pyrrolidone (NMP) for 30 min. This caused air bubbles to rise to the
surface. It is presumed that capillary forces of the solvent (NMP)
draw the individual CNTs closer to each other and as a result the
cross section of the CNT thread is reduced. The process resulted in
the fibres having an approximate 30% diameter reduction with
respect to the original, while densifications in acetone produced a
24% diameter reduction [32].
The afore-mentioned solvent densification typically improved
the thread's overall tensile strength and made it easier to handle
when adding knots with varying tightness. However, after solvent
densification, the CNT thread still maintained radial compressibility,
which was utilized in determining the tightness of the knot,