Could I suggest you use a surfactant with a phenyl or better still a naphthyl group as the hydrophobic part, the pi-pi interactions will help adsorption.
Re the settling, even if colloidally stable you may get some settling simply due to gravity, the only way to overcome this would be to increase the viscosity of the continuous phase in some way, or by weakly aggregating the system and have a concentration such that the volume the CNT's want to settle to is actually larger than the volume available. I do not know whether any one has done this for CNT's I doubt it actually but its done routinely for spherical particles, such as non-drip paint.