Ultra-Permeable Single-Walled Carbon Nanotube Membranes with Exceptional Performance at Scale

Abstract

Enhanced fluid transport in single-walled carbon nanotubes (SWCNTs) promises to enable major advancements in many membrane applications, from efficient water purification to next-generation protective garments. Practical realization of these advancements is hampered by the challenges of fabricating large-area, defect-free membranes containing a high density of open, small diameter SWCNT pores. Here, large-scale (≈60 cm²) nanocomposite membranes comprising of an ultrahigh density (1.89 × 10¹² tubes cm^-2) of 1.7 nm SWCNTs as sole transport pathways are demonstrated. Complete opening of all conducting nanotubes in the composite enables unprecedented accuracy in quantifying the enhancement of pressure-driven transport for both gases (>290× Knudsen prediction) and liquids (6100× no-slip Hagen-Poiseuille prediction). Achieved water permeances (>200 L m^-2 h^-1 bar^-1) greatly exceed those of state-of-the-art commercial nano- and ultrafiltration membranes of similar pore size. Fabricated membranes reject nanometer-sized molecules, permit fractionation of dyes from concentrated salt solutions, and exhibit excellent chemical resistance. Altogether, these SWCNT membranes offer new opportunities for energy-efficient nano- and ultrafiltration processes in chemically demanding environments.

Keywords: chemical resistance; enhancement factors; high‐density SWCNTs; large‐area CNT membranes; nanofiltration.

Figure 1

Fully opened SWCNT membranes. a) Pictures of small 1 cm² and large 60 cm² membranes before mounting (left) with SEM imaging of the composite membrane cross‐section (right). b) CNT diameter distribution determined by TEM with an average of 1.7 ± 0.7 nm (inner diameter 1.4 ± 0.7 nm). Inset: Representative TEM image of an SWCNT. c) Nitrogen permeance (P_N2) increase during membrane opening with oxygen plasma etching. Line used to guide the eye. d) Nitrogen permeance measured at different applied pressures. e) Measured ideal gas selectivity (gas X relative to nitrogen, X/N₂) of CNT membranes compared to predictions based on Knudsen diffusion. Parity line (y = x) included for comparison. Error bars represent the standard deviation of two to three membranes and are smaller than the data point in most cases. f) UV‐vis spectra of Direct Blue 71 (DB71) and 5 nm PEG‐coated gold nanoparticle (Au‐NP) solutions used in filtration experiments with a 1 cm² membrane. Feed and permeate spectra are shown in solid and dashed lines, respectively. Inset: Pictures of feed (F, colored) and permeate (P, clear) solutions.

Figure 2

Transport through fully opened SWCNT membranes. a) Nitrogen permeance (P_N2) for 1 and 60 cm² SWCNT membranes compared to prediction based on Knudsen diffusion. Error bars represent the standard deviation of three membranes. b) Reported gas enhancement factors for CNT membranes with 7.7 (),^{[ 37 ]} 7 (,^{[ 17 ]} ),^{[ 18 ]} 1.6 (),^{[ 3 ]} and 1.2 nm ()^{[ 23 ]} tube diameter compared to the 60 cm² SWCNT membranes with 1.7 nm () tube diameter from this work. Error bars represent the standard deviation of four membranes, both small and large. c) Pure water permeance (P_H2O) for 1 and 60 cm² SWCNT membranes compared to prediction based on no‐slip Hagen–Poiseuille flow. Error bars represent the standard deviation of three membranes. d) Pure water enhancement factors in CNTs from literature experiments (data points)^{[ 3} , ¹⁵ , ²¹ , ³⁸ , ³⁹ , ^{40 ]} and simulations (lines)^{[ 11} , ¹³ , ⁴¹ , ⁴² , ^{43 ]} compared to our results for a 60 cm² SWCNT membrane. Error bar for this work represents the standard deviation of four membranes (both small and large) and is smaller than the data point.

Figure 3

Properties of SWCNT membranes and comparison to literature. a) 3D parameter space comprising CNT diameter, number density, and membrane transport area for reported vertically aligned and partially aligned CNT membranes compared to our large‐area membranes.^{[ 4} , ¹⁶ , ¹⁷ , ¹⁸ , ²³ , ³³ , ³⁹ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² , ^{53 ]} b) 2D projections of CNT number density and tube diameter as a function of membrane transport area. c) Image of 60 cm² membrane rejecting over 99% of the DB71 probe analyte. d) Water permeance reported for commercial nanofiltration (NF, ) and ultrafiltration (UF, ) membranes compared to this work (). Commercial product information is reported in Tables S6 and S7 in the Supporting Information.

Figure 4

Separation performance of fully open SWCNT membranes. a) NaCl and Na₂SO₄ rejection at 1 bar applied pressure for a 1 cm² SWCNT membrane. Lines represent the rejection predicted by the Donnan model. b) Single component (1 m NaCl or 10 × 10⁻⁶ m Rose Bengal) and mixed feed (10 × 10⁻⁶ m Rose Bengal in 1 m NaCl) separation performance for 1 and 60 cm² SWCNT membranes. c) Rejection of 1 × 10⁻³ m NaCl, 0.33 × 10⁻³ m Na₂SO₄, and 10 × 10⁻⁶ m Rose Bengal feed solutions at various applied pressures (0.14, 0.28, 0.55, 0.69, and 1 bar) for a 1 cm² SWCNT membrane. Lines used to guide the eye. d) Rose Bengal rejection (left axis) and solution permeance (right axis) for a 1 cm² membrane before and after 2 h exposure to a 2000 ppm bleach cleaning solution.