Page 174 - Carbon Nanotube Fibres and Yarns

Carbon nanotube yarn structures and properties 165

In rubbing densification [29], roller pressure, roller speed, and yarn
tension could influence the density and properties of rub-densified CNT
yarns. The yarn tension was controlled by the speed ratio between the yarn
take-up (collection) speed and the circumferential speed of the rubbing
rollers. A high yarn take-up rate had a positive effect on the yarn specific
modulus but its effect on the yarn tenacity was not significant. On the
other hand, the yarn-specific modulus was increased by 56% when the yarn
take-up ratio was increased from 1.00 (zero tensioning) to 1.03, which ap-
peared to be the optimum tension level.

7.2.5 Post-spinning treatments
Wang et al. [33] reported multifold improvement of yarn strength by re-
peatedly rolling a solvent-densified yarn. The resulting flat yarn showed a
3
density estimated to be around 1.3–1.8 g/cm and an average strength of
4.34 GPa, which can be converted to a tenacity between 2.4 and 3.3 N/tex.
Zhang et al. [13] reported that “post-spin twisting” could dramatically
increase yarn breaking stress from 0.81 to 1.91 GPa. During the post-spin
twisting, a weight was hung on one end of the CNT yarn to provide tension
in the axial direction while the other end of the yarn was twisted. The tension
provided by the weight would increase the CNT packing density within the
yarn. The yarn diameter reduced from 4 to 3 μm after the post-spin twisting,
which represented a 44% reduction in yarn cross-sectional area. Similar bene-
fit of post-spinning twisting was also reported by Ghemes et al. [54].
Liu et al. [9] reported an 83% increase of yarn strength (from 0.6 to
1.1 GPa) by additional solvent-densifying treatment of twisted yarns. Less
than a half of the increase could be attributed to the increase of yarn break-
ing force and the rest to the yarn diameter reduction (increase of yarn
density). Li et al. [61] used different solvents to densify low twist yarns
(twist angle < 12 degrees) spun from CNT forests. In Fig. 7.20A, the sol-
vents are sorted according to their dipole moments. The strength of the
resulting yarns did not show a clear dependence on polarity. The fibers den-
sified by highly polar but nonvolatile solvents (slow evaporation), such as N,
N-dimethylformamide, dimethyl sulfoxide, and N-methyl-2-pyrrolidone,
were 100–200 MPa stronger than those densified by volatile solvents, such
as ethanol and acetone (fast evaporation). The fast vaporizing ethanol and
acetone result in uniformly distributed pores in the yarn while low vola-
tility solvents draw CNT in localized high-density areas with larger pores
between them. Cho et al. [32] reported that wet stretching with NMP and
CSA infiltration could significantly improve the specific strength of twistless