1 INTRODUCTION
As is known, hydrogen is one of the green and environmentally friendly clean fuels in the world, which has the potential to replace traditional fossil fuels for carbon peaking and carbon neutrality goals. In industry, a large number of hydrogen-containing mixed gas, mainly H2 and CO2, are produced through the steam reforming process of natural gas, thus subsequent purification is required to obtain high-purity hydrogen.1 Pressure swing adsorption (PSA) and low-temperature distillation are usually applied in industrial gas separation and purification, but both are energy-intensive processes. According to relevant research reports, energy consumption during the separation stage is extremely high, accounting for up to 85% of total energy costs.2 To reduce energy consumption and cost, researchers and engineers have turned their attention to membrane separation technology, which typically requires less investment and operating costs than mainstream alternative technologies such as PSA and cryogenic distillation, which is also more economical and energy-efficient.3-7However, the most challenging issue in membrane separation is efficiently preparing membranes with high selectivity/permeability and stability.
Although the polymeric membranes are easy to prepare at low cost, their separation performance is usually limited by the Robson Bound, where the gas permeability and selectivity have a trade-off phenomenon, thus the corresponding H2/CO2 separation performance is far away from the requirement for industrialization.8-13 Many other membranes composed of metal organic membranes (MOFs), covalent organic frameworks (COFs), zeolite molecular sieves (Zeolites), or other types of materials are also restricted from complex and time-consuming preparation process, difficulty in forming a continuous defect-free membrane layer, high cost, etc.14-30 In this decade, a kind of lamellar membrane based on 2D materials has attracted increasing attention, which shows the potential to break through the performance upper bound of traditional membranes.31 Various 2D nanosheet membranes constructed by graphene,32 graphene oxide,4,33-37 molybdenum sulfide,38,39 MXenes,40-43 2D zeolites,44,45 2D MOF,46,47 and 2D COF,48 etc. have been studied for gas separation, pervaporation and ion sieving, etc. In our previous work, the lamellar membranes based on MXene, a type of 2D transition metal carbides or carbonitrides prepared on the porous disk-shaped anodic aluminum oxide (AAO) substrate, showed good H2/CO2separation performance.42 In addition, Jin et al.43 prepared an ultra-thin (20 nm thick) MXene membrane on an AAO substrate, showing the performance of hydrogen permeability of 1584 GPU with H2/CO2selectivity of 27. Moreover, the research group proposed an external force-driven assembly approach (EFDA) to prepare a series of GO membranes on flaky α-Al2O3 substrates, which exhibited good molecular sieving performance with H2/CO2 selectivity of 30 and hydrogen permeability of 1000 GPU.35 On brittle AAO substrates, a g-C3N4-GO membrane was developed with outstanding hydrogen permeation capability (hydrogen permeance: 645 GPU, H2/CO2 selectivity: 39).36 Furthermore, Lai et al.38prepared GO-MoS2 hybrid membranes via a vacuum filtering approach on flaky AAO substrates, which exhibited hydrogen permeance of 857 GPU with H2/CO2 selectivity of 44. H2/CO2 selectivity of 30 and hydrogen permeance of 70 GPU were observed in self-crosslinked MXene membranes produced by filtering on yttria-stabilized zirconia hollow fiber.41
Although many 2D material-based membranes have been reported in the field of gas separation, the disk-shaped AAO, α-Al2O3, or some other fragile substrates were often chosen as the substrates for 2D nanosheet assembly in most research, which are not suitable for practical application due to the limited membrane area, difficulty of sealing, high-cost, and brittleness of the substrates. On the contrary, tubular membranes are more commonly used in industrial applications, which present the advantages of a relatively small footprint with large effective membrane area per unit volume, higher packing density, ease to seal for gastight, convenient to replace or repair any one from the entire membrane modules if necessary, etc. There are a few works that utilized tubular ceramic substrates for membrane construction, which are still too fragile in practical application for gas separation.40,41 Till now, seldom study has been reported to prepare 2D nanosheet membranes using cheap commercial porous substrates, whose pore size is usually large up to 20 μm with big curvature, because membrane defects are more likely to occur during the preparation process, resulting to a poor gas separation performance.41 Actually, one commercial stainless steel tube with macro-size pores (~2 μm) is a promising substrate. Since the porous stainless steel tube exhibits good corrosion resistance, high-temperature resistance, enough mechanical strength, and good welding performance compared with other materials, which is convenient to integrate with the equipment of different unit operations for H2 production and purification. Besides the substrate, the membrane preparation method also determines whether it could be utilized for scale-up production in industrialization, where a fast and efficient membrane assembly route with acceptable repeatability is urgently required.
In this work, a series of tubular MXene membranes are successfully constructed on commercially available macroporous stainless steel via electrophoresis, which is considered to be a simple preparation process with high efficiency and good membrane uniformity. The entire membrane preparation, including substrate modification, can be finished in 2~10 minutes. The membranes had exceptional gas separation performance with H2/CO2selectivity of 55 and hydrogen permeability of 1290 GPU. The separation temperature, water vapor and relative humidity, preparation process repeatability and long-term stability of the tubular MXene membranes have been investigated in detail, which is helpful and valuable for future industrial applications. Furthermore, this study would help promote the amplification of tubular MXene membranes and provide some experience for other 2D materials to be closer to actual industrialization.