Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1226-0088(Print)
ISSN : 2288-7253(Online)
Membrane Journal Vol.31 No.3 pp.184-199

Recent Advances in Covalent Triazine Framework based Separation Membranes

Esther Kim, Rajkumar Patel†
Energy and Environmental Science and Engineering (EESE), Integrated Science and Engineering Division (ISED), Underwood International College, Yonsei University, 85 Songdogwahak-ro, Yeonsu-gu, Incheon 21983, South Korea
Corresponding author(e-mail:;
May 26, 2021 ; June 27, 2021 ; June 28, 2021


As a branch of covalent organic frameworks (COF), covalent triazine frameworks (CTF) are inherently porous structures composed of networks of repeating hexagonal triazine rings fabricated via the ionothermal trimerization reaction. They also contain plenty of nitrogen functional groups that increase affinity for some chemicals while rejecting others. Because of their tunable properties, many researchers have synthesized and tested CTFs for gas and liquid separation processes. Various studies of novel CTFs, mixed CTF composites, and CTF membranes have experimented for gas adsorption/ separation (e.g., CO2, C2H2, H2, etc.) and desalination. Some CTF studies have determined the limits and potentials through advanced computer simulations while subsequent experiments have tested CTFs for photocatalytic properties, suggesting recyclability for greater sustainability. In this review, the covalent triazine framework-based separation membrane is discussed.

공유결합 트리아진 구조체 기반 분리막의 최근 발전

김에 스더, 라즈쿠마 파텔†
연세대학교 언더우드국제대학 융합과학공학부 에너지환경과학공학


공유결합 유기 구조체(COF)의 한 가지로서, 공유결합 트리아진 구조체(CTF)는 이온 열 삼량 체화 반응을 통해 제조된 반복되는 육각형 트리아진 고리의 네트워크로 구성되어 본질적으로 다공성 구조를 가진다. 또한 일부 화학 물질에 대 한 친화성을 높이고 다른 화학 물질을 배제하는 많은 질소 작용기를 포함한다. 조절 가능한 특성 때문에 많은 연구자들이 기 체 및 액체 분리 공정을 위한 CTF의 소재를 합성하고 테스트했다. 새로운 CTF, 혼합 CTF 복합재 및 CTF 멤브레인에 대한 다양한 연구가 기체흡착, 기체분리(예 : CO2, C2H2, H2 등) 및 담수화에 대해 연구되었다. 일부 CTF 연구는 고급 컴퓨터 시뮬 레이션을 통해 한계와 잠재력을 결정했으며 후속 실험에서는 광촉매 특성에 대한 CTF를 테스트하여 더 큰 지속 가능성을 위 한 재활용 가능성을 제안했다. 이 총설에서는 공유결합 트리아진 구조체 기반 분리막에 대해 설명할 예정이다.

    1. Introduction

    In an environment with an increasing need for fluid purification, such as CO2 capture and desalination, adsorption and/or separation technologies are highly demanded. To tackle micro pollution, organic frameworks are vastly investigated as they are generally known to be very stable under various pressures, chemicals, and temperatures. Their properties, such as porosity and surface energy, are also highly tunable which makes them useful in specific applications. Metal organic frameworks (MOF) are micro/nanoparticles with a metal core surrounded by a cage of organic linkers, making them highly porous and crystalline. The downside to MOFs,however, is the high material costs and their unknown effects in the open environment as an inorganic nanomaterial[ 1-7].

    Covalent organic frameworks (COF) are an alternative as they are cost-efficient and more environmentally friendly due to their fully organic nature; a specific branch of COFs is the covalent triazine framework (CTF). CTFs are characterized by their triazinering monomers that are repeated to form a highly porous structural network. Trimerization - the process of constructing hexagonal triazine rings - is generally initiated via an ionothermal reaction, where a Lewis acid catalyst (e.g., ZnCl2 and CF3SO3H) is added to the solution and heated. Not only are triazine rings mechanically strong channels that promote adsorption and permeability, but they also contain abundant nitrogen groups that encourage chemical selectivity. Inspired, scientists have proposed novel CTFs to test for adsorption and others have synthesized new CTF-modified membranes for separation processes. All in all, the following studies have discovered advantageous properties of CTFs and have advanced these frameworks as promising technology for large-scale purification systems[8-18].

    This article reviews recent CTF studies on gas separation (namely CO2 capture), desalination, and a few others. For each study, we discuss the synthesis of the CTF, the synthesis of the membrane (for some studies), characterizations, and finally adsorption/separation performances. The general fabrication of the CTF membrane and separation process are explained in Fig. 1 and study results are summarized in Table 1.

    2. Covalent Triazine Framework Membranes

    2.1 Gas Separation Membrane

    Several CTFs were synthesized in a vacuum at 400°C for 48 h by an ionothermal reaction between aromatic nitriles and ZnCl2 catalyst. Such nitriles included tetrakis (4-cyanophenyl)ethylene, terephthalonitrile, tetrafluoroterephthalonitrile, 4,4’-biphenyldicarbonitrile, and 1,3,5- benzenetricarbonitrile (the nitriles are referred to as M, M1, M2, M3, and M4, respectively)[19]. Mixed-nitrile CTFs were prepared similarly, but by combining M with M’ during synthesis. Cross-polarization magic-anglespinning nuclear magnetic resonance spectroscopy (CPMAS- NMR) analyzed H/C and F/C in MM2 and suggested a copolymeric CTF. Brunauer-Emmett-Teller (BET) characterization obtained the surface area measurements, confirming the largest surface area for M-CTF and smallest for MM2-CTF. The CTFs were observed for CO2 sorption and separation at 273 or 293 K and 1 bar. Five nitrile linkers (M and M1-4) were investigated for the synthesis of CTFs. Accordingly, five single-nitrile CTFs were created as the controls (Mand M’-CTF). This study by Dey et al. introduced a mixed-linker assembly strategy for nitrile-based CTFs and was the first to synthesize four novel mixed-linker CTFs that combines M with one other nitrile linker into a single separation CTF (MM’-CTF). Under BET analysis, all mixed-nitrile CTFs had significantly higher surface areas than their single-nitrile counterparts, with the exception of MM3-CTF; likewise, CO2/N2 selectivity for MM1-, MM2-, and MM4-CTF were higher. Interestingly, M-CTF had the largest surface area of 2,235 m2 g-1 - compared to 1,800, 1,360, 1,884, and 1,407 m2 g-1 for MM1-4, respectively. Although M-CTF had the largest surface area, results showed that MM1-, MM2-, and MM4-CTFs had the highest CO2 uptake capacities (i.e., 3.68, 4.70, and 3.40 mmol g-1, respectively), in which M-CTF was fourth highest with 3.26 mmol g-1. Note that, despite having the smallest surface area among mixed-nitrile CTFs in this study, MM2-CTF showed the second highest CO2 uptake capacity among reported CTFs synthesized at 400°C. The reason for these enhancements was most likely due to the higher micropore volume for CO2 accessibility created by defluorination carbonization during synthesis. It was concluded that, in CO2 storage at low pressures, surface area may not be a significant factor compared to CO2-accessible micropore volume.

    CTF-1 was synthesized by mixing terephthalonitrile and ZnCl2 via the ionothermal process at 400°C for 48 h in a vacuum. Polysulfone (PSF) was dissolved and mixed with different loadings of CTF-1 (i.e., 8, 16, or 24 wt%) to fabricate the novel mixed-matrix membrane (MMM)[20]. Scanning electron microscopy (SEM) characterization on the surface of the MMMs showed uniform dispersion of CTF-1 in the MMMs (Fig. 2). Accelerated Surface Area and Porosimetry (ASAP) analyzed the membrane surface, showing strong interfacial contact between the CTF-1 particles and PSF. The fabricated CTF-1@PSF MMMs and pure PSF membrane (0 wt%) was tested for gas separation at 25 °C and 3 bar. According to Dey et al., CTF membranes were synthesized to study the separation of CO2 from O2, N2, and CH4 gases. For good gas storage and separation CTF membranes, the parameters used were high surface area, low density, thermal and chemical stability, and plentiful nitrogen functionalities. Hence, CTF-1 and PSF were selected to form mixed-matrix membranes. Although O2, N2, and CH4 permeabilities showed no significant increase with the incorporation of CTF-1, CO2 permeability showed noticeable improvement. For instance, the 24 wt% MMM exhibited 5.4 barrer higher than the pure PSF membrane. In addition, CO2/N2 selectivity showed improvement with increasing loadings of CTF-1, the 24 wt% MMM exhibiting 26 compared to 23 from the pure PSF. Theoretical approaches, such as Maxwell models, have been successfully applied, which predicted that MMMs have the potential to yield about six times the permeability of pure PSF for CO2 and CH4 gases. Further analysis suggested that filler porosity has significant influence on the permeability of the membrane.

    The novel 2,6-bis(4-cyanophenyl)-1,5-dihydro-benzo [1,2-d:4,5-d’]diimidazole (BCDI) was synthesized by combining benzene-1,2,4,5-tetraamine tetrahydrochloride, 4-cyanobenzaldehyde, and NaHSO3 at 140°C for 24 h [21]. The prepared BCBDI was mixed with ZnCl2 in N2 atmosphere to undergo an ionothermal trimerization reaction at 550°C to form the novel CTF-DI. Thermogravimetric analysis (TGA) confirmed CTF-DI to have good thermal stability. BET characterization obtained surface areas that greatly varied across nine CTF-DIs that were synthesized under different conditions. CO2 capture and separations were tested at 273 K and 1 bar. Previous investigations on porous benzodiimidazole (BDI) based CTFs for gas separation found this CTF to be very inefficient because of strict conditions and unsatisfying permeability. BDI, however, was found to enhance the binding affinity between the CTF and CO2; therefore, an alternative CTF-DI synthesis methodology was proposed by Du et al. For the first time, BCBDI was created and used to synthesize the novel CTF-DI. The prepared CTF-DIs were characterized to have high surface areas up to 1,877 m2 g-1 and were tested for gas separation. Results indicated a high CO2 uptake of 89.2 cm3 g-1 and high adsorption heats of 52 kJ mol-1. Furthermore, selectivities were significant according to the ideal adsorbed solution theory (IAST): CO2/CH4 was up to 15, and CO2/N2 was a record showing up to 53. These selectivities were due to the abundant presences of active nitrogen species. Pore tunability of the CTF-DIs was also notable as adjusting catalyst ratios and synthesis temperatures was shown to greatly impact pore characteristics and surface area.

    A fluorine-based CTF, FCTF-1, was prepared via an ionothermal reaction between 2,3,5,6-tetrafluorotere- phthalonitrile (TFTPN) and ZnCl2 in a vacuum at 400°C for 40 h (Fig. 3). To synthesize a polymer of intrinsic microporosity (PIM), TFTPN, 5,5’,6,6’-tetrahydroxy- 3,3,3’,3’-tetramethyl-1,1’-spirobisindane (TTSBI), and K2CO3 were dissolved at 70°C for 40 h to create PIM-1[22]. The PIM-1 and FCTF-1 were then combined through solvent evaporation to fabricate the novel PIM-1@FCT-1 MMM. TGA investigated the thermal stability of FCTF-1, which indicated a relatively high thermal stability, most likely due to high bond energy of the triazine rings (Fig. 4). SEM characterization of the membranes analyzed cross-sectional images, which showed that adding FCTF-1 increased roughness and were dispersed uniformly within the matrix due to its organic nature. Indeed, some aggregates began forming at higher loadings, such as at 5 and 10 wt%, but no cracks or defects were detected. The MMMs were tested for CO2 separation at 35°C and 1 atm. PIMs are known to have excellent permeability with the drawback of only moderate selectivity in gas separations. This study by Jiang et al. aimed to improve gas selectivity, specifically CO2/CH4, by fabricating a novel FCTF-1 MMM. The FCTF-1 was combined with and PIM-1 matrix, which showed strong interfacial compatibility between the two phases because of the organic nature of the CTF. Moreover, fluorine’s naturally attraction to polar CO2 over nonpolar CH4 - in addition to the CTF’s microporous triazine-ring structure (Fig. 5) - increased CO2 sorption through the membrane while increasing CH4 rejection. The PIM-1@FCTF-1 MMM synthesized with 2 wt% filler loading exhibited the best overall performance, exhibiting 7300 barrer for CO2 permeability and 16.6 selectivity (Fig. 6). In addition, all fabricated MMMs (1-10 wt%) showed better permeability and selectivity than the pure PIM-1 membrane, which was 5800 barrer and 11.5, respectively. Even though the addition of FCTF-1 blocked many gas transport paths, the good CO2 solubility of the fillers encouraged permeance through the membrane. Further experiments on propene/propane gas separation proved successful with the novel MMMs, indicating potential in other small molecule separations.

    The hexene-CTFs were synthesized by the ionothermal process of trans-3-hexenedinitrile and ZnCl2 at 400 or 500°C for 48 h on a vacuum. Higher temperatures reorganized the nitrile groups into larger frameworks of triazine rings[23]. Powder X-ray diffraction (PXRD) characterization observed the crystallinity of the CTFs, which showed increased amorphous nature and surface area with more ZnCl2. This was supported by SEM and transmission electron microscopy (TEM), which visualized the crystalline clusters that were more prominent in CTFs that were synthesized with less ZnCl2 and lower temperature. The CTFs were tested for acetylene (C2H2), ethylene (C2H4), and CO2 separation from CH4 at 0 or 25°C under pressures up to 1 bar. Conventional hydrocarbon separation methods are energy-intensive, which hinders efficiency in large industries. As a low-energy alternative, Krishnaraj et al. introduced the hexene-CTF, being the first to have utilized a family of aliphatic olefin in the synthesis of a separation CTF. Various synthesis temperatures and conditions have shown to alter texture and porosity, indicating tunable properties (e.g., surface areas are higher at 500°C than 400°C due to defects causing mesopores). Additionally, microscopic images showed significant structural variations, showing the largest crystalline clusters for the hexene-CTF_400_1 (notation: CTF_synthesis temperature_ZnCl2 ratio) while showing spaghetti-like lattice planes for hexene- CTF_500_10. The structural significance was supported by test results of C2H2, C2H4, and CO2 adsorptions, which suggested that pore structure had greater influence in performance compared to surface area. In general, at low pressures, a higher-ordered CTF showed better adsorption of C2H2 and C2H4 despite having a smaller surface area, most likely due to the increased availability of existing functional groups. The highest adsorption was of C2H2 uptake with 3.85 mmol g-1 at 0°C and 1 bar, and both selectivities for C2H2 and C2H4 from methane (CH4) were also good due to the CTFs’ better interactions with molecules with double bonds compared to single bonds. CO2 adsorption and selectivity, however, were only average among reported CTFs, explained by hexene-CTF’s fewer interactive nitrogen groups compared to others.

    In the synthesis of CTF-PO71, Pigment Orange-71 (PO71) and ZnCl2 were combined at 400°C for 40 h under N2 atmosphere. The ionothermal trimerization process formed triazine rings to finalize the framework[ 24]. Fourier transform infrared (FTIR) characterization indicated complete polymerization and structural integrity of PO71 monomer during trimerization. Thermal analysis of the CTF-PO71 showed excellent chemical and thermal stability in air and N2, a highly demanded property in temperature-intensive industries. The novel CTF was tested for C2H2/C2H4 separation at 273 or 298 K and 100 kPa (1 bar). Pure C2H4 is in high demand as it is a key ingredient in the most commonly manufactured plastic: polyethylene (PE). A novel CTF, the CTF-PO71 created by Lu et al., was a simpler and cost-effective alternative to traditional industrial methods when it comes to removing difficult traces of C2H2 from C2H4. As it was prepared from an organic pigment molecule, the CTF-PO71 possessed functional sites and electrostatic potentials at the pore surface that were more compatible with C2H2 than C2H4. This enabled a strong interaction between the CTF and C2H2 and therefore effectively captured remnants of C2H2 while rejecting C2H4 molecules. For example, CTF-PO71 demonstrated one of the highest C2H2 adsorptions reported: 104 and 74 cm3 g-1 at 273 and 298 K, respectively. Ethylene separation from other gases were also tested at 1 bar: CO2 adsorptions were only 78.1 and 48.5 cm3 g-1 for 273 and 298 K, respectively (similarly, CH4 did not show any promising results). The reduction in performance for smaller gas molecules suggest that the properties of the pore surface were significant factors in addition to the pore shape and size. Being the first to report a successful C2H2/C2H4 separation method using porous organic polymer, this study is useful when considering large-scale gas separation processes.

    Covalent triazine piperazine polymer (CTPP) was synthesized by mixing piperazine, o-dichlorobenzene, N-diisopropylethylamine, and cyanuric chloride (C3Cl3N3) at 180°C for 3 days in normal atmosphere. To synthesize a pure poly(ether-block-amide) (PEBAX® 1657) matrix, a solution made from PEBAX powder was casted and dried[25]. The CTPP/PEBAX MMMs were prepared similarly with the addition of CTPP loadings in the PEBAX powder solution. TGA measurements of the MMM showed thermal stability up to 400°C. FTIR characterization confirmed the presence of triazine rings, although indications were not clear due to overlap with the PEBAX. The analysis, however, also showed increase in hydrogen bonding within the matrix when CTPP was added to the PEBAX. All membranes were tested for CO2 separation at 293K and 3 bar. As PEBAX is a matrix that exhibits high CO2 permeability but moderate N2 and CH4 selectivities, this study by Thankamony et al. incorporated CTPP into the matrix to create a new enhanced MMM for advanced CO2 separation processes. CTPP’s organic structure paired well with the PEBAX polymer due to strong hydrogen bonding, so the novel CTPP/PEBAX MMM was mechanically stable. In addition, the prepared CTPP was characterized and shown to be stable both chemically and thermally, which are favorable characteristics for any separation membrane. Further characterization showed increased chain rigidity and shrinkage in CO2 transport channels in the PEBAX when CTPP was added. Interestingly, this drawback was completely compensated for with the natural CO2-philic properties of the CTPP. In the separation of CO2/N2 and CO2/CH4, the CTPP/PEBAX MMM with only 0.025 wt% filler loading significantly improved membrane performance compared to the pure PEBAX membrane. More specifically, CO2 permeability increased from 53 to 73 barrer, and selectivities increased by 28 and 8 for N2 and CH4, respectively. The enhancement was explained by CTPP’s high porosity, surface area, and nitrogen content; overall permeability was greatly enhanced because the nitrogen attracts CO2 over N2 and CH4, which increased pure CO2 adsorption.

    In the computational synthesis of CTF-0, the trimerization reaction of carbonitriles was applied to create C3N3H3 triazine rings. These triazine building units were formed into two-dimensional (2D) sheets, which were then exfoliated into ultrathin nanosheets to be stacked into a porous monolayer membrane[26]. Density functional theory (DFT) calculations of the membrane described gas adsorption and permeation, indicating chemical stability with various gases; this is advantageous over other 2D membranes that are highly reactive, such as graphene. The Arrhenius equation was applied to estimate the membrane selectivities for He and H2, which showed a decrease in selectivity with increasing temperatures. COFs have intrinsic tunable pore properties that are beneficial for gas separation applications. By layering ultrathin 2D COF nanosheets, micropores can be uniformly distributed in a separation membrane. Inspired by the potential in 2D frameworks, Wang et al. simulated a novel CTF-0 monolayer membrane for He and H2 gas separation (Fig. 7). Diffusion energy barrier calculations on the CTF-0 membrane for eight different gases showed a direct correlation between membrane diffusion and electron density overlap, i.e., a higher density overlap corresponds to higher diffusion energy barrier. Existing as the two lightest el ements, He and H2 had the lowest diffusion barrier, exhibiting excellent permeance through the CTF-0 membrane at middling temperatures (Fig. 8). For instance, He and H2 exceeded the industrially acceptable permeance standard (10-10 mol m-2 s-2 Pa-1) at 300 and 355 K, respectively, improving with increasing temperatures. Contrastingly, permeance was significantly lower for Ne, CO, CO2, N2, CH4, Ar and gases - never reaching industrial standards up to 500 K - confirming the membrane’s exceptional selectivity for He and H2 over other gases (Fig. 9). When compared to most ultrathin membranes, the proposed CTF-0 membrane was superior in selectivities for He and H2 (Fig. 10).

    CTF-1 was synthesized by the ionothermal reaction between 1,4-dicyanobenzene (DCB) and ZnCl2 at 673 K in a vacuum. The resulting bulk powder was exfoliated into ultrathin nanosheets, which were mixed with graphene oxide (GO) nanosheets. Isopore cellulose acetate was used as the support membrane[27]. The GO-assisted CTF-1 membrane was fabricated by mixing and layering the two kinds of nanosheets thinly onto the substrate. It is a difficult to restack exfoliated 2D-COF without defects. FTIR characterization confirmed both the presence of triazine rings and undisrupted structures of CTF-1 nanosheets during membrane synthesis. Energy dispersive spectroscopy (EDS) illustrated the even dispersion of nitrogen atoms of the CTF in the membrane, which proved the continuous layering of the membrane. The prepared membranes were tested for hydrogen adsorption and separation at 298 K and 3.0 bar. 2D framework layering techniques to create gas separation membranes have shown to exhibit a high selectivity due to uniform pore structures. For example, GO and CTFs can be formed into 2D nanosheets, which, when stacked into membranes, are proven to be effective for separation processes. By stacking CTF-1 and GO nanosheets, Ying et al. fabricated a unique GO-assisted CTF-1 membrane. Because the GO contained abundant functional groups, it acted as a magnet between the CTF-1 layers, strongly adhering the nanosheets onto the cellulose support. (Note that insufficient amounts of GO were shown to produce defects due to poor adhesion). The membrane properties were tunable, meaning, different thicknesses (different amounts of nanosheets) adjusted narrowness of interlayer passages, which in turn affected selectivity and permeability. H2/CO2 separation using the 100 nm membrane indicated the highest recorded H2 permeance of 1.7 x10-6 mol m-2 s-1 Pa-1, and the 290 nm membrane indicated good selectivity of 17.4. Indeed, with increasing thickness, permeance decreased and selectivity increased. Nonetheless, the performance surpassed the Robeson’s 2008 upper bound standard for gas separation, deeming the GO-assisted CTF-1 membrane useful for practical H2 purification applications.

    To synthesize the triazine-framework membrane (TFM-1), 4,4’-biphenyldicarbonitrile was combined with trifluoromethanesulfonic acid (CF3SO3H) by trimerization and solvent evaporation at 100°C for just over 1.5 h under N2 atmosphere[28]. Two more membranes were fabricated using similar procedures, but with 1,4-dicyanobenzene and 2,7-naphthalenedicarbonitrile (TFM-2 and -3, respectively) (Fig. 11). TGA measurements of the membrane showed high thermal stability that parred with conventionally-prepared CTF membranes (Fig. 12). Energy dispersive X-ray spectroscopy (EDX) indicated negligible acid defects, which was contrasted to conventional CTF membranes that are commonly synthesized with acidic ion remnants. The prepared TFMs were then tested for CO2 adsorption and separation at 273 K and 1 bar. CTFs are commonly synthesized via the ionothermal trimerization reaction catalyzed by a Lewis acid, such as ZnCl2. Although results are usually impressive, the preparation conditions of using such acids are extreme - high pressures and temperatures of at least 400°C - making CTFs both difficult and dangerous to create. An alternative preparation methodology was proposed by Zhu et al., where the Lewis acid is replaced by the fluorescent superacid, CF3SO3H. This strong protic catalyst was found to achieve trimerization of three nitrile groups at low temperatures of around 100°C. In terms or experimental results, the TFM-1 exhibited an ideal CO2/N2 selectivity (29±2) and CO2 permeability (518±25 barrer). When compared to COF-102, which had a 5-times greater surface area, the TFM-1 exhibited a better CO2 uptake of 1.3 mmol g-1. The high CO2/N2 selectivity was explained by the nitrogen content in the triazine rings that attracted polarizable CO2 more than N2. Furthermore, the yellow TFM-1 membrane may be applicable in optoelectronics as the membrane possessed strong electron capture, emitting blue fluorescent under UV light (Fig. 13).

    2.2 Desalination membrane

    Theoretical 2D CTF models were designed using classical molecular dynamics (MD) simulations. The membrane was created under Lennard-Jones parameters, water molecules followed the SPC/E model, and salt ions followed parameters proposed by Joung et al.[29]. The REPEAT algorithm based off of DFT calculations derived charges of membrane atoms. By using an approach by Cohen-Tanugi et al., the mechanical integrity of the membranes under reverse osmosis (RO) pressure was tested, measuring stress as a function of PA pore radius. Desalination membranes are widely studied as precious water sources are increasingly polluted and require utmost attention. Pure polymer substrates, such as polyamide (PA), are known to exhibit high salt rejection but poor water permeability. To improve permeability, many studies have implemented porous nano frameworks into the substrate, such as carbon nanotubes, zeolites, and MOFs. As there is a lack of COF studies for liquid separations, this computational study by Lin et al. was the first to consider 2D CTF membranes for RO desalination. Simulations applying MD were employed under previously proposed water- and salt-molecule parameters, resulting in the formation of four ultrathin CTF membrane models. Theoretical desalination results for the CTF membranes showed complete salt rejection and 2-3 times the water permeability of PA-based membranes; these enhancements were greatly influenced by pore properties, especially pore size. Factors that were not considered in the experiments include membrane flexibility, thermal fluctuations, and deformation caused by pressure. However, easy tunability and structural integrity of the 2D CTF materials is very promising in the field of water purification.

    The synthesis of NENP-1 CTF involved mixing melamine, dimethyl sulfoxide (DMSO), and C3Cl3N3 at 150°C for 4 days in a vacuum. Poly(ether sulfone) (PES) was used as the support membrane[30]. The novel NENP-1-PSA/PES membrane was fabricated via the interfacial polymerization of poly(allylamine hydrochloride) (PAH) aqueous phase - combined with NENP-1 CTF nanosheets - and 1,3-benzenedisulfonyl dichloride (BDSC) organic phase to form a CTF-doped polysulfonamide (PSA) layer on the PES substrate (Fig. 14). Cross-polarization magic-angle-spinning (CP-MAS) analysis verified the presence of triazine rings, indicat ing the successful trimerization of NENP-1. FTIR characterized NENP-1 surface area and pore size, showing pathways larger than water molecules for smoother permeation (Fig. 15). The CTF membranes were tested for desalination at room temperature and 0.5 MPa. Few studies have ventured into COF technologies in liquid separations, and even fewer have explored COF use in purifying metal-contaminated waters. Hence, Wang et al. recently synthesized a 2D CTF, NENP-1, and fabricated a novel NENP-1-PSA/PES membrane for the nanofiltration of acidic wastewaters. The NENP-1 nanosheets were chosen for their hydrophilic properties, positive charge surfaces, and appropriate pore size to simultaneously improve water permeability and inorganic salt rejection (Fig. 16). PSA was chosen as the matrix as it proved to have good interfacial interactions and stability with NENP-1. The 0.1 w/v% CTF-infused membrane exhibited 75.4 L m-2 h-1 for pure water permeability and 93.3% MgCl2 rejection (Fig. 17). When compared to an un-doped PSA membrane, the NENP-1-PSA/PES membrane was shown to increase permeability by 149% and rejection by 4.6%. Furthermore, various tests confirmed the intrinsic structural stability of the novel membrane, which was complemented by excellent long-term acid stability and performance in pressurized water separation processes. Although research for COF/CTF membranes in desalination procedures is just beginning, there is great potential for such technological advancements in the near future.

    2.3 Others

    Basic CTFs were synthesized by the ionothermal reaction of 4,4’-biphenyldicarbonitrile and CF3SO3H in an ice bath for 1.5 h, then 100°C for 20 min under N2 atmosphere. These CTFs were then exfoliated with sulfuric acid into CTF nanosheets (CTFNS). Bulk graphitic carbon nitride (GCN) was amine-functionalized before also being exfoliated into nanosheets (CNNS)[31]. To synthesize the CTFNS/CNNS composite, a fixed amount of CNNS was mixed with different loadings of CTFNS (1, 5, 10, and 20%) and gone through the electrostatic self-assembly method (i.e., annealed at 200°C for 2 h (2°C min-1) in normal atmosphere). TEM characterization visualized and confirmed the porous 2D structures of both CTFNS and CNNS in the composite. X-ray photoelectron spectrum (XPS) analysis indicated a stronger intensity of N and O in the composite compared to pure CTFNS and CNNS, respectively, suggesting the presence of both nanomaterials in the composite. Photocatalysis with the CTFNS/CNNS composite was tested under the irradiation of simulated solar light. The increase in the use of sulfonamide antibiotics have significantly contributed to environmental pollution, which calls for methods to remove it, especially in wastewaters. Photocatalysis is a cost-effective and green method for degrading sulfonamide antibiotics, and Cao et al. proposed the novel CTFNS/CNNS composite, which is a metal-free 2D/2D heterojunction photocatalyst. Individually, GCNs and CTFs are semiconductors that exhibit weak photocatalytic activity, which can be moderately improved by changing their morphologies to 2D nanosheet. When combined, the binary Langmuir-Hinshelwood model explained that these 2D carbon-based materials had significant electrostatic interfacial contact, which enhanced the transfer and separation of excited-state electrons. In terms of results, the 5 wt% CTFNS/CNNS showed the highest photocatalytic activity, decomposing 10 ppm of sulfamethazine (SMT) in 180 min (95.8% degradation efficiency). Further experiments suggested that the composite’s photocatalytic performance can improve in the presence of chlorine ions as they encourage the production of reactive species (e.g., *O2- and *OH). In addition, an experiment with high performance liquid chromatography-mass spectrometer identified several intermediate products during the degradation of SMT, such as sulfanilamide and sulfanilic acid, which helped propose possible pathways of SMT degradation. The novel CTFNS/CNNS composite showed great potential for 2D/2D heterojunction carbon materials as photocatalysts in water purification processes.

    The 2D-CTF-1 nanosheets were synthesized by ionothermal cyclotrimerization of CF3SO3H with DCB and dichloromethane (CH2Cl2) solution on micro-interfaces at 100°C for and dried in a vacuum. Anodic aluminum oxide (AAO) was used as the substrate[32]. The nanosheets were evenly dispersed in dimethylformamide (DMF) before being coated (with different thicknesses) onto AAO substrate via gas-driven pressure filtration. The CTF and membrane morphology was characterized by SEM, where a uniform carbon matrix is visualized on the surface, and the cross-sectional layering of each nanosheet in the membrane can be seen. XRD analyzed the crystallinity of the membranes, indicating an interlayer spacing between the 2D-CTF-1 to be 3.3 Å, which was similar to graphene. The prepared membranes were tested for desalination from molecules or ions at room temperature (25°C) and increasing pressures. Separation membranes using 2D building blocks is highly researched as they exhibit exceptional tunability. In this study, Li et al. synthesized the novel 2D-CTF-1 nanosheet and tested its desalination abilities when fabricated into a membrane. (GO-doped membranes are also fabricated for comparison). The 2D-CTF-1 nanosheets were shown to possess high crystallinity and were successfully, and thinly, dispersed onto the substrate. In addition to the 2D properties, the 2D-CTF-1 nanosheets were very porous due to their triazine structures, or “skeleton pores”, which en hanced permeability for water while maintaining high selectivity for larger molecules and ions. For instance, the novel membrane exhibited excellent water permeance of 141.5 L m-2 h-1 under 1 bar of pressure and rejection above 95%. Even under increasing pressures up to 5 bar, rejections remained consistently higher than 95%, and permeance linearly increased as the highly crystalline 2D-CTF-1 nanosheets encouraged membrane rigidity and stability. Furthermore, the exact nano size of the hexagonal skeleton pores (i.e., 1.39 nm) was extremely beneficial for molecule/ion sieving. In addition to its advanced separation performance, the 2D-CTF-1 membrane was shown to be stable in aqueous environments and higher pressures, deeming it largely advantageous in the separation field.

    The micro-interface method was deployed to synthesize 2D-CTF-1 nanosheets, i.e., trimerization of DCB and CH2Cl2 solution with CF3SO3H at 100°C for 1 h. CdS/CTF-1 nanocomposite sheets were synthesized by heating a solution of CTF-1, cadmium acetate, and DMSO at 180°C for 3 h[33]. Using AAO as the substrate, membranes were synthesized by mixing the three prepared samples with DMF before evenly coating onto the substrate via gas-driven filtration method. Atomic force microscopy (AFM) characterization confirmed the successful fabrication of CTF-1 nanosheets, showing two layers with the combined thickness of about 5 nm. Electron-spin paramagnetic resonance (EPR) spectrometer was used to detect free radicals, which found characteristic peaks for *O2- and *OH after irradiation. The prepared membranes were tested for dye separation and photocatalytic sterilization under 0.1 MPa of pressure. Fouling of separation membranes can cause issues of sustainability and efficiency. A green and recyclable CTF separation membrane was proposed by Li et al., in which fouled membranes were purified via photo-assisted catalysis. More specifically, the membrane was fabricated with a novel composite consisting of photocatalytic CdS quantum dots (QD) loaded on 2D-CTF-1 nanosheets. Water transportation channels expanded when planting CdS QDs onto the CTF-1 nanosheets. As a result, the CdS/CTF-1 composite exhibited a high water flux of at least 179 L m-2 h-1 while keeping high dye rejections of above 94%. The membrane also showed impressive photocatalytic performance when assisted with H2O2, recovering more than 95% of the permeability under multiple cycles of sunlight irradiation. The improved photo-cleaning function is explained by the heterojunction between the 0D (0 dimensional) quantum dots and 2D nanosheets, which recombined photo-generated electron-hole pairs. The CdS-doped CTF-1 membrane is a potential breakthrough for further study on more sustainable separation membranes.

    COF particles were synthesized by creating ethanol solutions of melamine (Me), benzidine (Bd), and 1,3,5- triformylphloroglucinol (Tp) before collecting the COFs (TpBd and TpBdMe) via centrifugation. PSF was used as the substrate[34]. To fabricate the COF membranes, an aqueous solution of Bd and Me on the substrate was reacted with Tp organic solution (n-hexane) by interfacial polymerization. AFM confirmed an increase in membrane surface roughness when COFs layers were formed, most likely due to more cross- linking and rougher network structures. Water contact angle (WCA) observed the hydrophilicity of the membrane, which showed an improvement from pure PSF. The prepared membranes were tested for dye separation at 2.0 bar. Due to their advanced crystalline imide structures, COFs are highly desirable ingredients in separation membranes. A unique blend of monomers (i.e., Me, Bd, and Tp) was presented in this study by Wang et al. for the purpose of synthesizing a COF composite membrane. Two composite membranes, TpBd/PSF and TpBdMe/PSF, were shown to be very hydrophilic due to an increase in surface roughness compared to pure PSF. For instance, TpBd/PSF had high water permeability of 33.6 L m-2 h-2 bar-1 and TpBdMe/PSF was two times greater with 62.2 L m-2 h-2 bar-1 of permeability. The superior permeability of TpBdMe/PSF compared to TpBd/PSF was due to the introduction of benzene rings and amino groups from Me, which greatly enhanced hydrophilicity of the membrane. Moreover, the introduction of Me significantly modified the molecular structure, which consequently increased crystallinity and pore aperture. Dye rejection was also superior for both membranes, showing more than 99% rejection for Congo red. The membranes were highly sustainable as they showed good stability (chemically and thermally) and anti-fouling properties. This research explored imine-linked COF membranes and showed the promising tunability of membrane properties by introducing amine monomers.

    The triazine-based MOF1 was synthesized by the solvothermal reaction between a mixture of 5,5’,5’’- ((1,3,5-triazine-2,4,6-triyl)tris(azanediyl))-tris(2-methylbenzoic acid)/cadmium nitrate hexahydrate and a solution of dimethylacetamide/methanol at 120°C for 72 h[35]. PXRD characterization measured the crystallinity of MOF1, which was in accordance with simulated patterns, indicating successful synthesis without defects. TGA measured the thermal stability of MOF1 and determined a maximum of 360°C before it began decomposing. The ion adsorption performance and Cu2+ selectivity was observed via ICP-MS experimental analysis after filtration through a DMF solution of various metal nitrates. Excessive copper pollution is a potential hazard to life forms in the form of disabilities or birth defects. According to a study by Yousaf et al., adsorption of Cu2+ ions is an effective method to remove copper from the environment, and the novel MOF1 - which is triazine-based - is a great candidate. MOFs are inherently porous and stable, and with the incorporation of nitrogen-abundant sites, can improve selectivity for Cu2+ over other ions. In detail, after dispersing MOF1 in a solution of various metal ions, the MOF1 crystal was extracted and analyzed. Results indicated a high Cu2+ adsorption capacity of 50 mg g-1, which is the same performance as in a pure solution of only copper. In other words, there was no decrease in adsorption capacity for Cu2+. Column chromatography using these bright metal-ion solutions was also deployed to test MOF1’s separation properties. All mixtures of Cu2+ with one other metal ion, such as Co2+, were completely separated; that is, the copper-adsorbed MOF1 remained at the top of the column whereas the rest were easily drained (using DMF as the eluting agent). Similar adsorption tests were done with organic dyes, where Methylamine blue (MB+) was adsorbed much quicker than Acridine orange (AO-), suggesting a greater influence of ionic selectivity for cationic dyes rather than size. Finally, the triazine-based MOF1 was tested for encapsulating luminescent species, which, when combined with green terbium (Tb3+) and red Europium (Eu3+), the blue MOF1 could be tuned to emit quality white light. Not only did MOF1 show excellent copper ion electivity, but it also showed its potential as a luminescent material.

    3. Conclusions

    Microporous frameworks are advanced technologies that are crucial for efficient purification of gas and liquid systems by separation. CTFs are a branch of COFs that have recently caught attention for their facile and cost-efficient synthesis procedures as well as their excellent separation performances. Various microscopies that characterize morphologies, such as SEM and AFM, have shown the successful fabrication of defectfree triazine networks. Other characterization procedures have measured specific properties of the CTFs: generally, BET measured specific surface areas, TGA indicated thermal stability, and XRD confirmed crystallinity. In terms of performance, CTFs have shown consistently remarkable performance, improving upon substrates in the form of composite membranes and even improving upon itself by manipulation of nano-properties (e.g., exfoliation into 2D nanosheets). Additionally, some studies have investigated CTF’s potential in the field of photocatalysis, which resulted in discovering green methods to efficiently recycle fouled membranes via sun irradiation. In conclusion, covalent triazine frameworks are easily tunable organic materials with abundant nitrogen sites, which, in addition to their high porosity, have great potential in future advanced purification industries.



    Schematic of synthesis and separation process of CTF composite or membrane.


    Cross-section SEM images of 8 wt% (A), 16 wt% (B) and 24 wt% (C) of CTF-1@PSF MMM (Reproduced from Dey et al., 20, Copyright 2019, Frontiers in Chemistry).


    Synthesis Procedure of FCTF-1 (Reproduced with permission from Jiang et al., 22, Copyright 2020, American Chemical society).


    (a) FT-IR spectrum of FCTF-1; (b) XRD pattern of FCTF-1, indicating the amorphous structure of FCTF-1; (c) TEM image of FCTF-1; (d) TGA curve of FCTF-1, showing the high thermal stability (Reproduced with permission from Jiang et al., 22, Copyright 2020, American Chemical society).


    (a) Nitrogen adsorption and desorption isotherms of FCTF-1; (b) NLDFT pore size distribution of FCTF-1 (Reproduced with permission from Jiang et al., 22, Copyright 2020, American Chemical society).


    CO2 Permeability and CO2/CH4 selectivity of MMMs incorporating different content of FCTF-1 (Reproduced with permission from Jiang et al., 22, Copyright 2020, American Chemical society).


    Top view of the fully optimized 2 × 2 supercell of the monolayer CTF-0 (C, gray; N, blue; H, white) (Reproduced with permission from Wang et al., 26, Copyright 2016, American Chemical society).


    Energy profiles for (a) He and (b) H2 diffusion through the pore of monolayer CTF-0. (Insets) Corresponding configurations of IS, TS, and FS. For TS, both top and side views are given (Reproduced with permission from Wang et al., 26, Copyright 2016, American Chemical society).


    Electron density isosurfaces for (a) He, (b) Ne, (c) H2, (d) CO2, (e) Ar, (f) N2, (g) CO, and (h) CH4 passing through the pore of a monolayer CTF-0 membrane at the transition states. The isovalue is 0.12 e/Å3 (Reproduced with permission from Wang et al., 26, Copyright 2016, American Chemical society).


    Selectivities of the monolayer CTF-0 membrane for (a) He and (b) H2 over other gases as a function of temperature (Reproduced with permission from Wang et al., 26, Copyright 2016, American Chemical society).


    (A) Density functional theory (DFT)-optimized structures of the nitrile monomers used for the membrane synthesis. (B) Trimerization reaction of 4,4′-biphenyldicarbonitrile in CF3SO3H at 100°C. (C) Photograph of a directly synthesized sample of the transparent and flexible triazine-framework-based membrane TFM-1. Inside the circle is shown an “optimized” fragment of the hypothesized framework obtained using Materials Studio (Reproduced with permission from Zhu et al., 28, Copyright 2016, American Chemical society).


    (A) Comparison of the stabilities of (a) the 4,4′ -biphenyldicarbonitrile precursor (blue) and (b) the new synthesized membrane TFM-1 (red) under N2. (B) FTIR spectrum of TFM-1 prepared at 100°C. (C) 13C CP-MAS NMR spectrum of TFM-1. (D) Deconvolution of the N 1s peak in the XPS spectrum of TFM-1 (Reproduced with permission from Zhu et al., 28, Copyright 2012, American Chemical society).


    Photoluminescence spectrum of TFM-1. The inset shows the different colors of TFM-1 before and after excitation under UV light (Reproduced with permission from Zhu et al., 28, Copyright 2012, American Chemical society).


    Schematic preparation of NENP-1-PSA/PES nanocomposite membranes (Reproduced with permission from Wang et al., 30, Copyright 2020, American Chemical society).


    (a,b) TEM image of as-prepared NENP-1; (c) NENP-1 particle size distribution; (d) AFM image of as-prepared NENP-1; (e) height profile of NENP-1 (Reproduced with permission from Wang et al., 30, Copyright 2020, American Chemical society).


    Separation performance of NENP-1-PSA/PES membranes with different concentrations NENP-1 (Reproduced with permission from Wang et al., 30, Copyright 2020, American Chemical society).


    Schematic illustration of separation process for NENP-1-PSA/PES nanocomposite membrane (Reproduced with permission from Wang et al., 30, Copyright 2020, American Chemical society).


    Summary of CTF and CTF membranes


    1. M.A. Solomos, F.J. Claire, T.J. Kempa, “2D molecular crystal lattices: Advances in their synthesis, characterization, and application”, J. Mater. Chem. A, 7, 23537 (2019).
    2. J. Wang, S. Zhuang, “Covalent organic frameworks (COFs) for environmental applications”, Coord. Chem. Rev., 400, 213046 (2019).
    3. Z. Xia, Y. Zhao, S.B. Darling, “Covalent Organic Frameworks for Water Treatment”, Adv. Mater. Interfaces, 8, 2001507 (2021).
    4. J. Artz, “Covalent Triazine-based Frameworks - Tailor-made Catalysts and Catalyst Supports for Molecular and Nanoparticulate Species”, ChemCat- Chem, 10, 1753 (2018).
    5. B. Zheng, X. Lin, X. Zhang, D. Wu, K. Matyjaszewski, “Emerging Functional Porous Polymeric and Carbonaceous Materials for Environmental Treatment and Energy Storage”, Adv. Funct. Mater., 30, 1907006 (2020).
    6. A. Bhunia, I. Boldog, A. Möller, C. Janiak, “Highly stable nanoporous covalent triazine-based frameworks with an adamantane core for carbon dioxide sorption and separation”, J. Mater. Chem. A, 1, 14990 (2013).
    7. J. Zhu, S. Yuan, J. Wang, Y. Zhang, M. Tian, B. Van der Bruggen, “Microporous organic polymer- based membranes for ultrafast molecular separations”, Prog. Polym. Sci., 110, 101308 (2020).
    8. S. Dey, A. Bhunia, I. Boldog, C. Janiak, “A mixed-linker approach towards improving covalent triazine-based frameworks for CO2 capture and separation”, Microporous Mesoporous Mater., 241, 303 (2017).
    9. A. Boretti, S. Al-Zubaidy, M. Vaclavikova, M. Al-Abri, S. Castelletto, S. Mikhalovsky, “Outlook for graphene-based desalination membranes”, npj Clean Water, 1, 5 (2018).
    10. K. Yuan, C. Liu, J. Han, G. Yu, J. Wang, H. Duan, Z. Wang, X. Jian, “Phthalazinone structurebased covalent triazine frameworks and their gas adsorption and separation properties”, RSC Adv., 6, 12009 (2016).
    11. J. Safaei, P. Xiong, G. Wang, “Progress and prospects of two-dimensional materials for membranebased water desalination”, Mater. Today Adv., 8, 100108 (2020).
    12. J. Kim, M.R. Othman, “Research Trend on ZIF-8 Membranes for Propylene Separation”, Membr. J., 29, 67 (2019).
    13. J. H. Jo, W. S. Chi, “Review on Membrane Materials to Improve Plasticization Resistance for Gas Separations”, Membr. J., 30, 385 (2020).
    14. D. Kim, “Review on Zeolite MFI Membranes for Xylene Isomer Separation”, Membr. J., 29, 202 (2019).
    15. C. Yin, Z. Zhang, J. Zhou, Y. Wang, “Single- Layered Nanosheets of Covalent Triazine Frameworks (CTFs) by Mild Oxidation for Molecular-Sieving Membranes”, ACS Appl. Mater. Interfaces, 12, 18944 (2020).
    16. Z. Yang, W. Guo, S.M. Mahurin, S. Wang, H. Chen, L. Cheng, K. Jie, H.M. Meyer III, D.E. Jiang, G. Liu, W. Jin, I. Popovs, S. Dai, “Surpassing Robeson Upper Limit for CO2/N2 Separation with Fluorinated Carbon Molecular Sieve Membranes”, Chem, 6, 631 (2020).
    17. G.R. Xu, J.M. Xu, H.C. Su, X.Y. Liu, L. Lu, H.L. Zhao, H.J. Feng, R. Das, “Two-dimensional (2D) nanoporous membranes with sub-nanopores in reverse osmosis desalination: Latest developments and future directions”, Desalination, 451, 18 (2019).
    18. C. Zhang, B.H. Wu, M.Q. Ma, Z. Wang, Z.K. Xu, “Ultrathin metal/covalent-organic framework membranes towards ultimate separation”, Chem. Soc. Rev., 48, 3811 (2019).
    19. S. Dey, A. Bhunia, H. Breitzke, P.B. Groszewicz, G. Buntkowsky, C. Janiak, “Two linkers are better than one: enhancing CO2 capture and separation with porous covalent triazine-based frameworks from mixed nitrile linkers”, J. Mater. Chem. A, 5, 3609 (2017).
    20. S. Dey, S. Bügel, S. Sorribas, A. Nuhnen, A. Bhunia, J. Coronas, C. Janiak, “Synthesis and Characterization of Covalent Triazine Framework CTF-1@Polysulfone Mixed Matrix Membranes and Their Gas Separation Studies”, Front. Chem., 7, 693 (2019).
    21. J. Du, Y. Cui, Y. Liu, R. Krishna, Y. Yu, S. Wang, C. Zhang, X. Song, Z. Liang, “Preparation of benzodiimidazole- containing covalent triazine frameworks for enhanced selective CO2 capture and separation”, Microporous Mesoporous Mater., 276, 213 (2019).
    22. H. Jiang, J. Zhang, T. Huang, J. Xue, Y. Ren, Z. Guo, H. Wang, L. Yang, Y. Yin, Z. Jiang, M.D. Guiver, “Mixed-Matrix Membranes with Covalent Triazine Framework Fillers in Polymers of Intrinsic Microporosity for CO2 Separations”, Ind. Eng. Chem. Res., 59, 5296 (2020).
    23. C. Krishnaraj, H.S. Jena, K. Leus, H.M. Freeman, L.G. Benning, P. Van Der Voort, “An aliphatic hexene-covalent triazine framework for selective acetylene/methane and ethylene/methane separation”, J. Mater. Chem. A, 7, 13188 (2019).
    24. Y. Lu, J. He, Y. Chen, H. Wang, Y. Zhao, Y. Han, Y. Ding, “Effective Acetylene/Ethylene Separation at Ambient Conditions by a Pigment-Based Covalent- Triazine Framework”, Macromol. Rapid Commun., 39, 1700468 (2018).
    25. R.L. Thankamony, X. Li, S.K. Das, M.M. Ostwal, Z. Lai, “Porous covalent triazine piperazine polymer (CTPP)/PEBAX mixed matrix membranes for CO2/N2 and CO2/CH4 separations”, J. Membr. Sci., 591, 117348 (2019).
    26. Y. Wang, J. Li, Q. Yang, C. Zhong, “Two-Dimensional Covalent Triazine Framework Membrane for Helium Separation and Hydrogen Purification”, ACS Appl. Mater. Interfaces, 8, 8694 (2016).
    27. Y. Ying, D. Liu, J. Ma, M. Tong, W. Zhang, H. Huang, Q. Yang, C. Zhong, “A GO-assisted method for the preparation of ultrathin covalent organic framework membranes for gas separation”, J. Mater. Chem. A, 4, 13444 (2016).
    28. X. Zhu, C. Tian, S.M. Mahurin, S.H. Chai, C. Wang, S. Brown, G.M. Veith, H. Luo, H. Liu, S. Dai, “A superacid-catalyzed synthesis of porous membranes based on triazine frameworks for CO2 separation”, J. ACS, 134, 10478 (2012).
    29. L.-C. Lin, J. Choi, J.C. Grossman, “Two-dimensional covalent triazine framework as an ultrathin- film nanoporous membrane for desalination”, Chem. Commun., 51, 14921 (2015).
    30. H. Wang, H. Wang, H. Jiang, A. Sheng, Z. Wei, Y. Li, C. Wu, H. Li, “Positively Charged Polysulfonamide Nanocomposite Membranes Incorporating Hydrophilic Triazine-Structured COFs for Highly Efficient Nanofiltration”, ACS Appl. Nano Mater., 3, 9329 (2020).
    31. S. Cao, Y. Zhang, N. He, J. Wang, H. Chen, F. Jiang, “Metal-free 2D/2D heterojunction of covalent triazine-based frameworks/graphitic carbon nitride with enhanced interfacial charge separation for highly efficient photocatalytic elimination of antibiotic pollutants”, J. Hazard. Mater., 391, 122204 (2020).
    32. G. Li, W. Wang, Q. Fang, F. Liu, “Covalent triazine frameworks membrane with highly ordered skeleton nanopores for robust and precise molecule/ ion separation”, J. Membr. Sci., 595, 117525 (2020).
    33. G. Li, J. Ye, Y. Shen, Q. Fang, F. Liu, “Covalent triazine frameworks composite membrane (CdS/ CTF-1) with enhanced photocatalytic in-situ cleaning and disinfection properties for sustainable separation”, Chem. Eng. J., 424, 127784 (2020).
    34. T. Wang, H. Wu, S. Zhao, W. Zhang, M. Tahir, Z. Wang, J. Wang, “Interfacial polymerized and pore-variable covalent organic framework composite membrane for dye separation”, Chem. Eng. J., 384, 123347 (2020).
    35. A. Yousaf, A.M. Arif, N. Xu, J. Zhou, C.Y. Sun, X.L. Wang, Z.M. Su, “A triazine-functionalized nanoporous metal-organic framework for the selective adsorption and chromatographic separation of transition metal ions and cationic dyes and white-light emission by Ln3+ ion encapsulation”, J. Mater. Chem. C, 7, 8860 (2019).