Fast diffusive transport through carbon nanotube pores

Molecular diffusion across semi-permeable biological membranes capitalizes on complex nanoscale transport mechanisms that enable high permeation rates and finely tuned control over the profile of permeating molecules. Synthetic materials have thus far been unable to match their biological counterparts. Overcoming this obstacle has the potential to significantly impact applications such as hemodialysis, drug delivery, and biomolecular purification. Fluid flow through carbon nanotubes (CNTs) has been demonstrated to have enormous transport rates under a pressure driving force, yet comparatively little attention has been given to concentration driven transport. While bulk/hindered diffusion for small molecules has been assumed in some studies on nm-wide CNTs, other simulations and NMR experiments support the idea that self-diffusion coefficients may be several times larger than in bulk. These large uncertainties in the magnitude of the diffusion rates through CNTs have hampered their full exploitation in nanofluidic devices and membrane platforms. Here, we fabricate membranes with a large, well-characterized number of single-walled CNTs which enable an accurate extrapolation of the per-pore diffusive flow rate. A series of novel and stringent control experiments rule out the possibility of defects at all scales, including those smaller than the pore openings. Once corrected for boundary layer resistance at the membrane/fluid interface, our measurements indicate that the transport diffusivity of small ions in 2-nm single-walled carbon nanotubes is approximately 10-30 times larger than in bulk. By employing a variety of permeating molecules, we are compiling a large dataset to better understand the flow enhancement dependence on the chemico-physical properties of permeating molecules and to demonstrate the benefit of this transport phenomenon for dialysis separations. These results shed further light into the unique transport properties of graphitic channels and enable a more accurate design of CNT-based fluidic systems for a broad range of applications.