Dry-spun carbon nanotube ultrafiltration membranes tailored by anti-viral metal oxide coatings for human coronavirus 229E capture in water

Abstract

Although waterborne virus removal may be achieved using separation membrane technologies, such technologies remain largely inefficient at generating virus-free effluents due to the lack of anti-viral reactivity of conventional membrane materials required to deactivating viruses. Here, a stepwise approach towards simultaneous filtration and disinfection of Human Coronavirus 229E (HCoV-229E) in water effluents, is proposed by engineering dry-spun ultrafiltration carbon nanotube (CNT) membranes, coated with anti-viral SnO₂ thin films via atomic layer deposition. The thickness and pore size of the engineered CNT membranes were fine-tuned by varying spinnable CNT sheets and their relative orientations on carbon nanofibre (CNF) porous supports to reach thicknesses less than 1 µm and pore size around 28 nm. The nanoscale SnO₂ coatings were found to further reduce the pore size down to ∼21 nm and provide more functional groups on the membrane surface to capture the viruses via size exclusion and electrostatic attractions. The synthesized CNT and SnO₂ coated CNT membranes were shown to attain a viral removal efficiency above 6.7 log₁₀ against HCoV-229E virus with fast water permeance up to ∼4 × 10³ and 3.5 × 10³ L.m^-2.h^-1.bar^-1, respectively. Such high performance was achieved by increasing the dry-spun CNT sheets up to 60 layers, orienting successive 30 CNT layers at 45°, and coating 40 nm SnO₂ on the synthesized membranes. The current study provides an efficient scalable fabrication scheme to engineer flexible ultrafiltration CNT-based membranes for cost-effective filtration and inactivation of waterborne viruses to outperform the state-of-the-art ultrafiltration membranes.

Keywords: Antiviral metal oxide coatings; Carbon nanotube membrane; Fast water permeation; Human Coronavirus 229E; Virus ultrafiltration.

Graphical abstract

Fig. 1

. (a-c) The SEM morphology of CNF support and multi-layer CNT membranes. (d) The mean pore size variation of CNT membranes with the number of dry-spun CNT layers. (e-g) SEM images of 30-layer CNT membranes oriented at different angles (0,45, and 90°). (h) The change in the CNT membrane pore size with the orientation angle between the dry-spun CNT layers.

Fig. 2

SEM and corresponding TEM images of SnO₂-CNT membranes with varied SnO₂ thicknesses: (a-b) 5 nm, (c-d) 25 nm, and (e-f) 40 nm. (g) The variation of SnO₂ thickness with ALD cycles. (h) The variation of SnO₂-CNT membrane pore size with the SnO₂ thickness.

Fig. 3

(a) The variation of SnO₂ weight percent with SnO₂ thickness. (b) XRD spectra of the synthesized membranes at different SnO₂ thicknesses. (c) The deconvoluted O 1 s spectra of 25 nm SnO₂ coated CNT membrane. (d) The deconvoluted Sn 3d spectra of 25 nm SnO₂ coated CNT membrane. (e) The variation of streaming zeta potential at different pH values from 2 to 11. (f) The change of membrane wettability with SnO₂ thicknesses.

Fig. 4

Water permeance performance of: (a) CNF support and multi-layer CNT membranes, (b) CNT membranes with different orientation angles between 30 CNT layers, and (c) 30-layer CNT membranes deposited with different SnO₂ thicknesses. LRV of HCoV-229E of: (d) CNT-CNF membranes with varied CNT layers, (e) 30-layer CNT membranes with different orientation angles, and (f) 30-layer CNT membranes deposited with different SnO₂ thicknesses.

Fig. 5

A schematic presentation of the virus rejection mechanisms by the different CNT-based membranes with: (a) Parallel multi-layer CNT sheets, (b) Oriented CNT sheets at different angles (0, 45, and 90°), and (c) SnO₂ coated CNT sheets.