Amino-decorated organosilica membranes for highly permeable CO₂ capture

Highlights

• Amino-decorated organosilica membrane was prepared via co-polymerization strategy.

• BTESA-A membranes obtained controlled pore size and enhanced CO₂ affinity.

• BTESA-A membranes featured promisingly high CO₂ capture performance.

Abstract

Membrane-based separation technologies are considered an effective process for the capture of CO₂ due to advantages such as high levels of energy efficiency with inexpensive levels of investment. It remains challenging, however, to develop a membrane with high levels of both permeance and selectivity with the ability to capture CO₂ in a highly efficient manner. Herein, we report amino-decorated organosilica membranes that are based on a facile and effective co-polymerization strategy that uses bis(triethoxysilyl)acetylene (BTESA) and (3-aminopropyl) triethoxysilane (APTES) precursors. This co-polymerization strategy simultaneously endows the resultant membranes with a controlled pore size and enhanced CO₂-philic properties that have significantly improved CO₂ separation performance. These composite membranes display CO₂ permeance that ranges from 2550–3230 GPU and a range for CO₂/N₂ selectivity of 31–42 in the separation of CO₂/N₂ mixtures, which outperforms most state-of-the-art membranes and exceeds the target for post-combustion CO₂ capture operations. A satisfying performance was achieved with a CO₂ permeance of 8 ✕ 10⁻⁷ mol m⁻² s⁻¹ Pa⁻¹ (2390 GPU) and CO₂/CH₄ selectivity of 70. The present study highlights an elegant decoration strategy for the production of membranes with ultrahigh CO₂ capture capacities, which can also be extended to other organosilica precursors for target-oriented separations by varying the bridged or side-chain groups.

Introduction

Excessive emissions of CO₂ have raised a series of environmental issues that are attracting intense levels of public scrutiny [[1], [2], [3]]. Compared with conventional separation processes, membrane-based separation techniques provide a promising strategy to achieve the goal of effective and energy-efficient CO₂ capture [[3], [4], [5]]. Hence, a variety of membranes that include organic [6], inorganic [7,8], and organic-inorganic hybrid varieties [9] have been used extensively for CO₂ separation. Among these, organosilica membranes with high hydrothermal stability and impressive molecular sieving properties have attracted a great deal of interest, and have shown great promise for use in CO₂ separation [[10], [11], [12], [13]]. However, it is intrinsically challenging to adjust the organosilica membrane structure with a pore size appropriate for high-performance CO₂ separation considering the close kinetic diameters of CO₂ (0.33 nm), N₂ (0.364 nm), and CH₄ (0.38 nm). In general, reinforcing the CO₂-philic performance and the molecular sieving properties of membranes has been deemed beneficial in promoting highly efficient CO₂ capture. The preferentially adsorbed CO₂ molecules on the membrane surface or network structures may hinder the permeation of other gases, as a result, the CO₂ separation performance are expected to be effectively enhanced. Therefore, we expected the combination of CO₂-philic and molecular-sieving properties to accelerate the exploitation of membranes for CO₂ separation.

A specific interaction can be generated between CO₂ molecules and amino groups [14], and (3-aminopropyl) triethoxysilane (APTES), which consists of aminopropyl groups, has been used both in the fabrication of sensors and as a modifying agent of adsorbents [15,16]. Consequently, with regard to membrane separation, the existence of amino groups in APTES is expected to accelerate the CO₂ separation properties of APTES-derived membranes. The resultant APTES membranes, however, have always been beset by the densified structures and the low utilization efficiency of amino groups, which seriously inhibits the application of highly permeable CO₂ membranes [11,14]. An APTES membrane reported by Xomeritakis et al. produced a low level of CO₂ permeance at 2.6 ✕ 10⁻⁸ mol m⁻² s⁻¹ Pa⁻¹ (77.7 GPU, 1 GPU equals to 3.348 ✕ 10⁻¹⁰ mol m⁻² s⁻¹ Pa⁻¹), albeit with a satisfyingly high CO₂/N₂ selectivity of 70 [14]. Consequently, APTES has been preferred as a modifier for possible advancements in CO₂ separation performance. Suzuki et al. fabricated a hybrid silica membrane via a chemical vapor deposition strategy that used tetraethylorthosilicate (TEOS) and APTES as precursors, and the resultant membrane returned a CO₂ permeance of 2.3 ✕ 10⁻⁷ mol m⁻² s⁻¹ Pa⁻¹ (687 GPU) with a CO₂/CH₄ selectivity of 40 [17]. In fact, the modification of membranes using organosilica has been a common strategy for elevating the gas separation performance for other types of inorganic membranes [18,19]. The APTES-functionalized ZIF-90 membrane exhibited improved H₂/X selectivities (X represents CO₂, CH₄, C₂H₆, C₃H₈) with a relatively high H₂ permeance of 2.9 × 10⁻⁷ mol m⁻² s⁻¹ Pa⁻¹ [18]. The incorporated APTES narrowed the membrane pore size and eliminated the formation of intercrystalline defects. A similar strategy was also applied to the SAPO-34-derived zeolite membrane [19]. The bis(triethoxysilyl)ethane (BTESE) precursor was used to functionalize the SAPO-34 membrane via a strategy of vacuum-assisted deposition to selectively patch the defects, and the as-synthesized composite membrane displayed improved CO₂/CH₄ selectivity.

In our previous study, an ultramicroporosity-tailored composite organosilica membrane was fabricated by the co-condensation of 4,6-bis(3-triethoxysilyl-1-propoxy)-1,3-pyrimidine (BTPP), which is CO₂-philic, and BTESE (or TEOS). This membrane produced an increase in CO₂ permeance (~2000 GPU) and a CO₂/N₂ selectivity that reached ~20 [12]. Generally, high permeance is preferred over high selectivity since the increased permeance through a membrane without a loss of selectivity has been recognized as more important in industrial scenarios. As we know, improving the selectivity to a level of more than 30 will not significantly decrease the operational costs, but enhancing CO₂ permeance would [1]. Hence, a higher possible level for CO₂ permeance with moderate CO₂/N₂ selectivity (20–30) would be the most desirable model for energy-efficient CO₂ capture [20,21]. The intrinsically dense structure of a TEOS network and the flexible ethane-bridges of a BTESE structure [22,23], however, still restrict the further advance of CO₂ separation performance for the resultant composite-membranes, particularly in the case of ultrahigh CO₂ permeance.

In the most recent work by our group, the network structure and permeation properties of organosilica membranes were effectively tailored via the control of bond angles [23]. By comparison with BTESE and bis(triethoxysilyl)ethylene (BTESEthy) membranes, bis(triethoxysilyl)acetylene (BTESA) membranes, feature rigid acetylene bridges and a more open and accessible network structure that attributed to the increased bond angles, which has been confirmed via FT-IR spectra and molecular simulation. BTESA membranes show great potential in C₃H₆/C₃H₈ separation due to a relatively larger and uniform pore size (around 0.52 nm) [24,25]. Nevertheless, this membrane has been difficult to use in either CO₂/N₂ or CO₂/CH₄ separations because of the small molecular sizes of CO₂ (0.33 nm), N₂ (0.364 nm), and CH₄ (0.38 nm). For instance, the selectivity of CO₂/N₂ (around 7 to 13) was much lower than the target separation performance (CO₂ permeance > 1000 GPU, CO₂/N₂ > 20) [1], even though the CO₂ permeance (~6500 GPU) was sufficiently high. Consequently, a design that could effectively increase the CO₂/N₂ selectivity while ensuring there is no significant sacrifice in the level of CO₂ permeance would be a worthwhile development for BTESA membranes.

Herein, APTES was introduced into a BTESA membrane matrix to tailor the microstructure and advance the CO₂ capture performance, as illustrated in Fig. 1. A plausible schematic illustration of the BTESA, BTESA-A, and APTES membrane structures appears in Fig. S1 of the supporting information. A co-condensation strategy, which is authoritative and advantageous in fine-tuning the physico-chemical structures of organosilica-based materials, was used to realize the target. The decoration of APTES in a BTESA matrix (referred to BTESA-A) enhanced the affinity of composite membranes to CO₂ molecules and simultaneously controlled the membrane pore size, which was confirmed by the measurements of N₂ sorption, CO₂ isosteric heat of adsorption, and gas permeation properties. In contrast to the pure BTESA and APTES membranes, the composite BTESA-A membranes achieved a high CO₂ separation performance that is unprecedented and superior to most of the membranes reported in the literature. The fabricated membranes and the composite strategy reported in the present study provide novel opportunities for high-performance membranes that can be used in critical separation processes.