All of the TIF-Ax compounds were initially synthesized using the solvothermal method (for detailed information, refer to the original papers). Based on the solvothermal method, diverse improved methods have been developed to serve various purposes.

TIF-A1 was originally synthesized by the solvothermal method using Zn(NO₃)₂∙6H₂O, adenine, and isonictinic acid (Hint) in N,N′-dimethylformamide (DMF) at 120 °C for 3 days with a high yield (88%) [22].

2.2. Metal Sources

Inspired by the aforementioned outcomes achieved through stirring and heating at 120 °C in a brief timeframe, and mindful of the potential for the reaction between nitrate and carboxylic acid to generate corrosive nitric acid, which might compromise the equipment’s integrity and influence the stability of the final MOFs, alternative metal sources were employed to investigate the feasibility of substituting Zn(NO₃)₂∙6H₂O. Similarly, the yields of TIF-A1 synthesized from ZnO and Zn(CH₃COO)₂∙2H₂O were 75.5% and 80.5%, respectively ^[63][67]. This shows that the metal source has good substitutability.

2.3. Upscale Synthesis

The potential for synthesizing a kilogram of TIF-A1 was further investigated using the aforementioned methods. Interestingly, through the scaling up of reactant quantities, a yield of over 800 g of TIF-A1 was achieved in a single heating process at 120 °C ^[63][67]. The resulting samples exhibited a similar crystallinity, morphology, and BET surface area to those obtained through the original solvothermal method.

3. Structure Diversity of TIF-Ax

TIF-A1 is a 4-connected metal–organic framework with Zn²⁺ as the metal center, adenine (ad) as the main ligand, and isonicotinic acid (Hint) as the auxiliary ligand. Both ligands bind to Zn²⁺ via N and carboxyl O on heterocycles. It was observed that TIF-A1~A3 has a different topology, which was regulated by changing the type of imidazole salts in the framework. 2-NH₂-TIF-A1 and 3-NH₂-TIF-A1 were prepared by functional modification, and their functional groups could effectively adjust the pore environment and pore size. Cd-ad-int (int= isonicotinic acid) changed the metal center to have different building units, thus adjusting the structural network. Zn-thp-nit (thp = theophylline, nit = nicotinic acid) changed the isonicotinic acid ligand into nicotinic acid, which changed the angle of connection between the ligand and the metal. In TIF-A4 ~ A8, monocarboxylic acid with a coordination angle of ca. 120° was selected to replace the Hint, and a dia topological structure and layered structure were obtained.

4. Special Properties of TIF-Ax

4.1. Solvent Stability

To realize the wide application of MOFs, they should first have good stability in harsh conditions, such as high temperature, high pressure, and acid and alkali environments. Interestingly, TIF-A1 can be stable in various organic solvents, such as water, N,N′-dimethylformamide (DMF), N,N′-diethylformamide (DEF), acetonitrile (CH₃CN), ethyl acetate, toluene, n-hexane, dichloromethane (DCM), chloroform, methanol (MeOH), ethanol (EtOH), isopropanol (IPA), and n-butanol (n-BuOH). Furthermore, TIF-A1 can also maintain the framework after being soaked in a solution of pH= 2–1 ^[63][67]. The exceptional stability of TIF-A1 makes it a promising candidate for further industrial applications.

4.2. Guest Selectivity

It is well known that solvent plays an important role in the structural variety of MOFs. Different solvent systems tend to produce different MOFs. However, it was found that the construction of TIF-A1 was independent of the solvents. TIF-A1 can accommodate amide solvents, including N,N′-dimethylformamide (DMF), ethyl urea (e-urea), N-methyl-1,2-pyrrolidone (NMP), and N,N′-dimethylacetamide (DMA). Interestingly, the crystallization process of TIF-A1 exhibited a strict selective trapping ability for these four solvent guests, and its selective order was DMF > e-urea > NMP > DMA (Scheme 2). This order of selection may be related to the size of the guest and the interaction between the host and guest. When the size of the guest is relatively small, the size of the guest is the main factor affecting the selection of guest molecules for TIF-A1. The four solvent guests are arranged in that order of molecular size: DMF < DMA < e-urea < NMP. DMF is very suitable to exist in the pores of TIF-A1. When the size of the guest is relatively large, TIF-A1 needs more energy to expand the pore to accommodate them. However, under this condition, the interaction between the host and the guest is dominant, and the energy needed to expand the pore will be partially offset. There is a strong N-H∙∙∙N hydrogen bond between e-urea and TIF-A1. There are non-classic C-H∙∙∙O and C-H∙∙∙N hydrogen bonds between NMP and TIF-A1. There is only a non-classic C-H∙∙∙O hydrogen bond between DMA and TIF-A1. Therefore, the selection order for TIF-A1 is e-urea > NMP > DMA. However, if the molecule is too large, it still depends mainly on the size of the guest. For example, TIF-A1 cannot trap 1,3-dimethyl-2-imidazolidinone (DMI) or N,N′-dimethylpropyleneurea (DMPU) ^[64][68].

4.3. Flexibility

As mentioned above, TIF-A1 showed a certain degree of contraction and expansion during crystallization to adapt to different guest molecules. Considering the porosity of TIF-A1, the flexibility of it was further explored by different gases ^[65][69]. The de-solvated form denoted as TIF-A1′ was obtained through methanol exchange. Unexpectedly, TIF-A1′ showed normal adsorption behavior for most common gases, such as N₂, H₂, and CO₂, while it only showed multistep adsorption for C₂H₂. With the combination of in situ single-crystal XRD and calculation simulation, the flexibility of TIF-A1 was derived from the strong interaction between acetylene and the uncoordinated carboxylate O atoms of the int ligand. The C₂H₂-induced flexibility of TIF-A1 is intrinsic and was not affected by the sample size or defects. However, 2-NH₂-TIF-A1 exhibited normal acetylene adsorption because the presence of a strong H-bond between the -NH₂ of ad and the uncoordinated carboxylate O atoms of 2-NH₂-int prevented C₂H₂ from approaching the active site and made the rotation of 2-NH₂-int ligand more complex.

5. Application

5.1. CO

₂

Separation

Considering the presence of abundant active sites on TIF-A1, its CO₂ absorption was first evaluated. After becoming fully activated, TIF-A1 exhibited high CO₂ uptake at 273 K and 1 bar (107 cm³/g) ^[64][68]. In contrast, the highest CO₂ adsorption in ZIFs at that time was ZIF-69, with a value of 70 cm³/g [22]. The high CO₂ adsorption capacity of TIF-A1 can be attributed to the synergistic effect of int and ad ligands. Because the -NH₂ group amino and pyrimidine N atoms in ad form an electron-rich system, they are potential CO₂ bond sites. As mentioned above, the uncoordinated carboxylate O atoms of int ligand can also be active sites. In addition, the specific pore size and environment of TIF-A1 also play important roles in CO₂ adsorption. Furthermore, the special interaction between TIF-A1 and CO₂ indicates that TIF-A1 has the potential ability to selectively adsorb CO₂, which makes it possible to apply TIF-A1 in the capture and separation of CO₂ in flue gas. The flue gas mainly contains CO₂ and N₂, which are similar in molecular size, and it is difficult to achieve the separation effect by the pure physical adsorption of porous materials. The N₂ adsorption of TIF-A1 was 4.7 cm³/g (273 K) and 1.4 cm³/g (298 K), which shows that TIF-A1 barely adsorbed N₂. Henry’s constant indicates that the adsorption selectivity of TIF-A1 for CO₂ over N₂ was 90 (273 K,1 bar) and 60 (298 K,1 bar), respectively ^[58][62]. The excellent separation selectivity of TIF-A1 was further demonstrated by a breakthrough experiment ^[63][67].

5.2. NH

₃

Adsorption

Due to the strong electronegativity of the N sites in TIF-A1, it can not only form an electron-rich system to attract CO₂ but also form a strong hydrogen bond with NH₃. Therefore, the NH₃ adsorption of 3-NH₂-TIF-A1 was studied. At 298 K and 1 bar, the maximum NH₃ adsorption capacities of 3-NH₂-TIF-A1 (obtained in DMA) was 9.8 mmol/g (obtained in DMA) and 7.1 mmol/g (obtained in DMF), which were higher than those of MOF-5 and ultrahigh porous MOF-177 ^[59][63]. 3-NH₂-TIF-A1 has abundant active sites, including an uncoordinated oxygen atom and -NH₂ group of 3-NH₂-int, uncoordinated N atoms and -NH₂ group of the ad ligand. Among them, the -NH₂ group on ad and the N of the pyrimidine ring at its para-position had lower E_ads values compared to those of the other four sites, which can provide more stable adsorption sites.

5.3. C

₂

Separation

In addition to CO₂, TIF-A1 also exhibited a good C₂ separation ability. In 2022, Ding and co-workers applied TIF-A1 to trap C₂H₂ and C₂H₆ simultaneously from the ternary mixture gas of C₂H₂/C₂H₄/C₂H₆ and performed the purification of C₂H₄ ^[66][70]. In the ternary mixture of C₂H₂/C₂H₄/C₂H₆, a strong electrostatic interaction occurred between C₂H₂ and the uncoordinated carboxyl O with high polarity in TIF-A1, and van der Waals (vdW) interaction occurred between C₂H₆ and the aromatic heterocycles with low polarity. TIF-A1 not only provides a strong binding site for the adsorption of C₂H₂ and C₂H₆, but also oblate C₂H₆ and linear C₂H₂ are very suitable for the spindle-shaped cage of TIF-A1, which can store the target molecule well. The results show that TIF-A1 separated 99.9% ethylene from C₂H₂/C₂H₄/C₂H₆ (1/10/89) at 298 K and 1 bar, and the yield was 1.43 mmol/g, representing the best purification capacity at that time.

5.4. CO

₂

Cycloaddition

In 2021, Wang and co-workers selected ZnX₂ (X = Cl, Br, I) instead of ZnNO₃ and selected a suitable solvent environment to synthesize a series of X-TIF-A1 with many halogen ions in the framework as catalysts for the CO₂ activation reaction ^[67][71]. Due to their potent nucleophilic nature, halogen ions induced the ring-opening of propylene oxide, facilitating the reaction between CO₂ and propylene oxide to generate propylene carbonate—a pivotal step in catalyzing CO₂ conversion. The outcomes demonstrated that the yield of the CO₂ activation reaction catalyzed by ZnI₂-ad-int-DMF exhibited an upward trend with increasing reaction temperature. Notably, the yield achieved an impressive 98.5% at 140 °C. Subsequent to three cycles of reuse, ZnI₂-ad-int-DMF experienced a moderate decline in catalytic activity. However, its framework remained stable despite exposure to high-temperature, high-pressure, and solvent conditions. Importantly, no discernible blocking phenomenon was observed as the iodine content decreased from 69 μmol/g to 51 μmol/g ^[67][71].

5.5. Heavy Metal Adsorption

In 2022, Ma and co-workers used chitosan as the TIF-A1 growth template and prepared TIF-A1/chitosan composite beads using a secondary growth method for Pb(II) adsorption in water ^[68][72]. Several small TIF-A1 crystals formed rod-shaped clusters and aggregated in the pores of chitosan. The average particle size of TIF-A1/chitosan was 0.2 cm. Such a large composite ball can effectively avoid the harm of pipeline blockage caused by the difficulty of separation and recovery of powdery MOF materials in practical applications. The polar groups -COO⁻, -OH, C=N, and -NH₂ in the TIF-A1/chitosan structure coordinate with Pb(II). The maximum adsorption of TIF-A1/chitosan for Pb(II) was 397.3 mg/g at 25 °C and pH = 6. Furthermore, in the mixed solution of multiple metal ions, TIF-A1/chitosan hardly adsorbed other ions, and the adsorption removal efficiency of Pb(II) was 99.17%. Especially when the concentration of Pb(II) was 100 ppb, the removal efficiency of trace Pb(II) was 99.95%, and the residual amount of Pb(II) met the international drinking water standards ^[68][72]. After five adsorption/desorption cycles, TIF-A1/chitosan could still maintain high adsorption performance, and the removal efficiency was more than 99%. Furthermore, the crystal structure of TIF-A1 was not destroyed, and TIF-A1 was still attached to the chitosan matrix. In conclusion, TIF-A1/chitosan is a promising adsorbent for the removal of trace Pb(II) in drinking water treatment because of its excellent performance and reusability.