Carbon-doped metal-oxide interface nanofilms for ultra-fast and precise separation of molecules
In industry, the separation of industrially relevant molecules, such as the pharmaceutical, petroleum, and chemical industries, requires the use of strong solvents at high temperatures. Most polymer membranes are prone to aging and/or collapse under these conditions, and non-polymer membranes are also problematic. So, we needed a material that could precisely control porosity, be easy to process, have a defect-free continuous membrane, and have excellent chemical, mechanical, and thermal stability. In addition, high permeability is also important considering the amount of solvent processed by industry. Reducing the thickness is a way to improve permeability, but it is easy to introduce defects. Professor Miao Yu of the State University of New York at Buffalo believes that increasing the density of nanopores can significantly improve permeability without having to thin the membrane to the limit. Their group reports a similar ultrafast interface process for the preparation of inorganic nanoporous carbon-doped metal oxide (CDTO) nanofilms that can be used for precise molecular separation. For a given pore size, these nanomembranes have a pore density (assuming the same curvature) 2 to 10 times higher than reported and commercially available organic solvent nanofiltration membranes, producing ultra-high solvent permeability even when they are thicker. Due to their excellent mechanical, chemical and thermal stability, CDTO nanomembranes have designable rigid nanopores for stable and efficient separation of organics over long periods of time under demanding conditions. The results were published in Science under the title "Carbon-doped metal oxide interfacial nanofilms for ultrafast and precise separation of molecules". The first author is Bratin Sengupta, and Qiaobei Dong is the co-author.
Preparation of porous nanofilms by interfacial reaction Inspired by molecular layer deposition (MLD) self-limiting reactions, the authors developed an interfacial reaction process to prepare a dense epidermal layer using titanium tetrachloride (TiCl4) and ethylene glycol (EG) as metal reactants and organic reactants, respectively (Figure 1A). This method generates OHF surface layers 2 to 3 orders of magnitude faster than layer-by-layer MLD processes. After carbon removal by heat treatment/calcification, nanoporous carbon-doped metal oxides (CDTOs) are generated (Figure 1A, IV). The authors optimized the manufacturing conditions for rapid and defect-free synthesis of organometallic hybrid films (OHFs). Liquid EG and TiCl4 vapor react at higher temperatures (150°C) to form the thinnest, defect-free OHF in the shortest amount of time. The authors modeled this hybrid material using a large-scale atomic/molecular massively parallel simulator (LAMMPS) to understand the formation of pores during calcination. Figure 1B shows the material formed after heat treatment in N2 and O2, respectively. Dense porosity is created throughout the material, and the resulting porous structure is highly dependent on the final carbon content (I and II in Figure 1B). The surface area, pore volume, and porosity of CDTO are highly correlated with the carbon retained in the titanium oxide network after calcination, i.e., "carbon doping" (Figure 1B, III). Carbon removal is responsible for pore formation and precise dimensional changes (Figure 1B). Controlling the carbon content also allows you to adjust surface properties; The higher the amount of carbon doped, the stronger the hydrophobicity. Figure 1D shows centimeter-scale OHF and CDTO nanomembranes with different compositions and properties on AAO and central control fibers (HF); The higher the carbon doping, the darker the color of the film: yellow for CDTO-Air and black for CDTO-N2. The SEM image (Figure 1E) shows a CDTO-Air selective layer approximately 30 nanometers thick on the AAO. The CDTO surface is smooth compared to porous substrates (Figure 1E).
Figure 1. Formation and characterization of CDTO nanomembranes Nanopores are rigid, stable, and selective, and the authors studied the transport of various organic solvents through CDTO nanomembranes and observed that permeability is 2 to 3 orders of magnitude higher compared to commercial OSN membranes. Solvents permeate more slowly in CDTO-N2 than CDTO-Air because of the smaller pores (higher carbon doping) produced when calcined in N2. The authors also prepared CDTO nanomembranes on central control fibers (HF) and tested their transport and separation efficiencies (Figure 2, B and C). Similar viscosity-dependent permeation was observed in CDTO nanofilms prepared on HF (Figure 2B). The results showed similar separation between CDTO-Air and CDTO-N2 prepared on AAO and HF supports (Figure 2C), which demonstrated that the CDTO holes were independent of the underlying support. CDTO-Air nanofilms were experimentally observed to stably separate rosy red in DMF at temperatures up to 140°C (Figure 2D).
Figure 2. Solvent permeation and dye rejection of CDTO nanomembranes precise pore size control by carbon doping For OSN, membranes with tunable nanopores are ideal with retained molecular weights between 200 and 1400 g/mol to meet the needs of various industrial processes. The authors employ two strategies to alter carbon doping: (i) changing the initial carbon content of dense OHF; (ii) Control of carbon removal by changing calcination conditions (gas environment/temperature). The resulting CDTO nanofilms have a precisely controlled intercepted molecular weight (Figure 2E) covering the entire OSN range. MWCO can be adjusted between 240 and 920 g mol-1 by varying the initial carbon content of OHF (Figure 2E). Another way to adjust the pore size is to control the removal of carbon (Figure 2E). Calcination at high temperatures or air removes more carbon from OHF, reducing carbon doping and creating larger pores. By increasing the calcination temperature in air from 250°C to 500°C, the intercepted molecular weight of CDTO nanofilms increased from 920 g mol-1 to >1000 g mol-1 (Figure 2E). Thus, CDTO has effective pore size tunable between 240 and 1400 g mol-1 (Figure 2E), while the step change in molecular weight cut-off is as small as 100 g mol-1. High-density nanopores produce ultrafast transport, and the authors believe that densely packed low-roundabout nanopores in CDTO are responsible for ultrafast solvent transport. High ε/τ (surface porosity/tortuosity) indicates that the high density of nanopores with low roundabouts facilitates transport (Figure 3A), eliminating the challenge of fabricating films "close to atomic" thickness. The authors calculated the ε/τ of CDTO nanofilms and reported OSN membranes using pure methanol flux and calculated effective pore size, and compared them as a function of molecular weight interception. The ε/τ of CDTO nanofilms generally increases with the increase of the retained molecular weight, with a maximum value of 0.175. The authors compared the OSN performance of CDTO nanomembranes with the retained molecular weight and pure methanol permeability of commercial and reported OSN membranes (Figure 3C). At the same cut-off molecular weight, the pure methanol permeability of CDTO nanofilms on AAO (high permeability support) is about 2 times higher than the reported maximum, which is clearly the result of high ε/τ. Even CDTO nanofilms on HF (Low Permeability Support) are comparable to the best films that have been reported.
Figure 3.Industrial separation of CDTO nanomembrane and OSN membrane under harsh conditions When synthesizing special chemicals such as small molecule drugs and pesticides, it is usually necessary to react under harsh solvent conditions of high temperature and high pressure. To verify the usefulness of the membranes, the authors selected CDTO membranes with suitable molecular weight cut-off and applied them to the production process of pesticides (Boscalid), including the separation of reactants, products, and homogeneous catalysts. The authors designed a two-stage membrane cascade system using two CDTO membranes with different molecular weight cut-offs, 940 and 300 g mol-1, to achieve (i) separation of catalyst from reactants and products, and (ii) separation of products from reactants (Figure 4A) in DMF at 90°C (Figure 4A). Verified by continuous cross-flow operation for 100 hours, CDTO demonstrated the ability to continuously reject catalyst and product, with significantly higher separation coefficients than Loose Film 1 (higher molecular weight interception) and Tight Film 2 (lower molecular weight interception), respectively, 65.9 (product/catalyst) and 17.4 (reactants/product), respectively, and remained stable in DMF at 90 °C (see Figure 4, B and C). These high separation factors enable membrane 1 to efficiently recover catalyst (loss less than 1%), extract 80% to 90% of reactants/products, and efficiently recover products (loss less than 5%) for recycling.
Figure 4. Summary of CDTO membrane separation under industrially relevant conditions: In summary, this study produced defect-free nanofilms by a fast interface reaction. CDTO nanomembranes exhibit extremely high stability in harsh solvents up to 140°C and have rigid nanopores that are precisely controlled across the entire OSN range, providing broad and precise pore adjustability for a single OSN membrane material. These films have a high ε/τ, possibly due to their excellent mechanical strength, enabling the formation of high-density and uniformly distributed nanopores. This allows CDTO nanomembranes to exhibit high permeability, even if they are not atomic-scale films. Due to their excellent stability, these membranes can be used in industrial processes that require harsh conditions. In addition, other suitable metal and organic reactants can be investigated in the future to prepare nanofilms generated at this interface for different separation applications. --Cellulose Recommended ---- Recommendation Number--
Source: Frontiers in Polymer Science