Chinese Chemical Letters  2025, Vol. 36 Issue (12): 110550   PDF    
A 1,3,5-triazine μ3-bridged neutral Cu(Ⅰ) framework with enhanced stability and CO2 capture selectivity
Feng-Fan Yanga, Yin-Kang Dinga, Lin-Kai Wua, Jiayue Tianb, Shuai Doub, Wenjing Wangc, Linfeng Lianga,d,*     
a Institute of Crystalline Materials, Shanxi University, Taiyuan 030006, China;
b School of Materials and Chemical Engineering, Zhengzhou University of Light Industry, Zhengzhou 450000, China;
c State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, China;
d Key Laboratory of Materials for Energy Conversion and Storage of Shanxi Province, Taiyuan 030006, China
Abstract: Compounds with 1,3,5-triazine ring such as melamine derivatives, known for their low cost and high stability, offer potential for affordable, stable metal organic framework (MOF) synthesis. However, reported such frameworks are facing issues like low stability and reduced porosity. To overcome such problems, we explored organic ligands with protonated nitrogen atoms on the 1,3,5-triazine ring, facilitated by introducing hydroxyl groups on carbon atoms of the triazine ring to construct MOFs. By using 5-azacytosine, we have successfully synthesized SXU-121, a neutral ultramicroporous MOF with eta-topology and high density of open Cu(Ⅰ) sites. SXU-121 features robust structural stability and significant CO2 adsorption capacity with high selectivity over N2 under ambient conditions. We believe SXU-121's development opens new avenues for creating a class of stable and low cost metal-triazine frameworks with potential diverse functionalities.
Keywords: 5-Azacytosine    Metal organic framework    Carbon dioxide capture    Crystal engineering    Gas adsorption and separation    

The three N atoms on 1,3,5-triazine ring possess excellent coordinating abilities and high C3 symmetry distribution (Fig. 1a), they offer a unique advantage in synthesizing highly symmetrical metal-organic frameworks (MOFs) like eta, pto, bor topology [13]. Compounds containing 1,3,5-triazine rings such as melamine and its derivatives, typically are very common industry products and have very low costs and high stability, making them highly promising for the synthesis of low-cost, high-stability MOFs [48]. However, current researches predominantly involves the direct coordination of neutral 1,3,5-triazine rings, which leads to MOFs with insufficient stability and large amount of counter ions which block the MOF pores, hindering the performance of related MOFs on many aspects, especially gas adsorption and separation [911]. Consequently, the challenge of synthesizing stable and porous MOFs containing 1,3,5-triazine rings remains a significant and meaningful research topic.

Download:
Fig. 1. Structural illustration of SXU-121. (a) The structure of 1,3,5-triazine. (b) The structure of selected 5-azacytosine ligand. (c) The coordination environment of Cu as 3-connected node. (d) The one-dimensional channels along the crystallographic c direction. (e) The 3-connected eta-topology of SXU-121. Cu atom, cyan; C atom, gray; O atom, red; N atom, blue; H atom, white.

Inspired by the ligands like imidazolates, triazolates or pyrazolates in the metal azolate frameworks [1215], which are known for their outstanding stability, we envision to select ligand with protonated nitrogen atoms on the 1,3,5-triazine ring to overcome the stability problems of 1,3,5-trizine based MOFs. By removing the protons and allowing the ring nitrogen atoms to coordinate with metal ions, we can directly obtain neutral 1,3,5-triazine based MOF materials while significantly enhancing their stability and keeping their permanent porosity. But how can we introduce protons onto the nitrogen atoms? It is found that the hydroxypyridine moiety possesses two tautomers, one of which allows the nitrogen atom to become protonated in the -keto tautomer form [7,1618]. Therefore, if hydroxyl groups can be introduced onto the carbon atoms of the 1,3,5-triazine ring, it should exhibit similar tautomers. Based on the above consideration, we focused our attention on a compound with a hydroxyl group on the triazine ring: 5-azacytosine (Fig. 1b). It is expected to hold the potential to construct stable, porous 1,3,5-trizine based target MOFs.

In this study, by reacting 5-azacytosine with copper ions under solvothermal conditions, we have synthesized a neutral ultramicroporous Cu(Ⅰ) framework SXU-121. This MOF showcases an eta-topology with permanent porosity and high density of Cu(Ⅰ) open sites. SXU-121 could retain its structural stability after 1 day in boiling water, demonstrating the feasibility of using 1,3,5-triazine ring derivatives as organic ligands to construct highly stable MOFs. Moreover, SXU-121 demonstrates high CO2 adsorption capacity and CO2/N2 adsorption selectivity at 298 K. This will open the avenue for constructing a class of metal-triazine framework materials with a wide range of functionalities.

Transparent, needle-shaped crystals of SXU-121 were obtained by solvothermal reaction of Cu(NO3)2·3H2O and 5-azacytosine in N,N-dimethylformamide (DMF) solvent at 120 ℃ for 3 days (30% yield, materials and synthetic procedures please refer Supporting information). Single-crystal X-ray diffraction analysis revealed that SXU-121 crystallizes in the trigonal crystal system with the space groups of P3221 (Table S1 in Supporting information). In SXU-121, each asymmetric unit includes one Cu+ and one deprotonated 5-azacytosine. One negatively charged N atom on the 1,3,5-triazine ring forms a charge-assisted coordination bond with the Cu(Ⅰ) ion, with a bond length of 1.989 Å. The remaining two N atoms form traditional Cu-N coordination bonds with lengths of 1.996 Å and 2.007 Å (Fig. 1c), respectively. These connections ultimately result in the formation of a neutral framework and 1D hexagonal channels with a pore diameter of 6.19 Å along the crystallographic c-axis (Fig. 1d), exhibiting a 3-connected eta-topological structure (Fig. 1e). The nitrogen atom of the amino group forms hydrogen bond with the oxygen on adjacent carbonyl group of the triazine ring (N-H⋯O, 2.97 Å, 166.14°) (Fig. S1 in Supporting information), which may be critical in inducing SXU-121 synthesis. The C═O double bond can be verified through the C═O bond length of 1.26 Å and the appearance of the absorption peak in Fourier transform infrared (FI-IR) spectroscopy at 1629 cm−1 (Fig. 2a), suggesting that 5-azacytosine ligand is presented as a carbonyl tautomer form. Similarly, the presence of this absorption peak is observed at the same position of SXU-121 (Fig. 2a), indicating that the C═O double bond is still retained in SXU-121. In the high-resolution X-ray photoelectron spectroscopy (XPS) of Cu 2p for SXU-121, the existence of the Cu+ with binging energies of 953.4 and 933.3 eV and the existence of the Cu2+ with binging energies of 956.1 and 935.4 eV suggest that a small amount of monovalent copper were oxidized to divalent copper during the crystal synthesis process (Fig. 2b). The Cu LMM Auger emission spectroscopy displays Cu LMM peaks with kinetic energy of 571.8 eV and 569.4 eV (Fig. 2c) giving further evidence that the presence of monovalent and bits of divalent copper in SXU-121. The pore size of SXU-121 is 6.19 Å as predicted by the Zeo++ software, classifying it within the ultramicroporous category [19]. The theoretical porosity of SXU-121 is 53.9% based on PLATON calculations with a probe radius of 1.8 Å [20].

Download:
Fig. 2. (a) FT-IR spectra of 5-azacytosine and SXU-121. (b) Cu 2p1/2 and 2p3/2 XPS spectra of SXU-121. (c) Cu LMM Auger electron spectroscopy of SXU-121. (d) The VTPXRD patterns of SXU-121. (e) The TG curve of SXU-121. (f) The PXRD patterns of SXU-121 after immersed in boiling water, basic environment in KOH solution (0.1 mol/L) and various organic solvents for 24 h.

The purity of bulk crystalline SXU-121 was confirmed through good accordance with the powder X-ray diffraction (PXRD) pattern simulated from the crystallographic information files (Fig. 2d). The thermal stability of SXU-121 was explored by thermogravimetric (TG) analysis combined with variable temperature powder X-ray diffraction (VTXRD). The TG curves revealed that the framework remains stable up to 280 ℃ as the temperature increases. SXU-121 underwent a two-step weight loss after 280 ℃ (Fig. 2e). The first step of weight loss occurred between 283 ℃ and 400 ℃ with continued exothermic reactions taking place in this interval, which is thought to be due to the polymerisation process of the organic ligand. The second step of weight loss occurred above 400 ℃ and is accompanied by a continuous exothermic reaction, which can be attributed in part to the decomposition of the polymer generated above. Furthermore, the consistent PXRD patterns at different temperatures confirm that SXU-121 can remain stable above 280 ℃ (Fig. 2d). To evaluate the chemical stability of SXU-121, the as-synthesized materials were treated under various conditions. After SXU-121 samples were placed in water, boiling water, 0.1 mol/L KOH solution, pH 2 solution and various organic solvents (n-hexane, ether, acetonitrile, CCl4) for 24 h, the chemical stability of SXU-121 was then explored by PXRD experiments. Under pH 2 acidic conditions, the structure tends to collapse, with noticeable changes observed in the PXRD pattern (Fig. S2 in Supporting information). Apart from this, the structure could be well retained after 24 h of treatment within other solutions or solvents except for 0.1 mol/L KOH solution with crystallinity decreased (Fig. 2f).

Prior to single component gas adsorption test, the SXU-121 sample was first activated. Approximately 100 mg of freshly synthesized SXU-121 crystals was placed in a 20 mL glass vial, repeatedly washed with dimethylformamide (DMF) under ultrasonication, and then immersed in 20 mL of ethanol for 72 h at ambient temperature. Throughout this period, the ethanol was refreshed every 12 h. Subsequently, the single crystals were filtered and subjected to vacuum drying treatment at 100 ℃ for 24 h. The PXRD curve of the vacuum dried SXU-121 as shown in Fig. 2f is in accordance with the simulated pattern, confirming its stability after activated. Following this, the vacuum-dried SXU-121 was degassed at 100 ℃ for 12 h and its N2 adsorption at 77 K was then evaluated. Although the theoretical pore size of SXU-121 (6.19 Å) is larger than the kinetic diameter of N2 molecule (3.8 Å), the N2 adsorption isotherm at 77 K demonstrated almost no N2 adsorption (Fig. 3a). This may be related to the fact that the open pore of SXU-121 has a weak affinity for N2 which has been reported before [21]. Through adsorption experiment of CO2 under conditions of 195 K, it was found that at a relative pressure of 1.0, the adsorption capacity reaches 70 cm3/g. The Brunauer-Emmett-Teller (BET) and Langmuir specific surface area were calculated to be 188 and 267 cm2/g, respectively, with a porosity of 9.2%. This performance is far from the results obtained from theoretical calculations. This may be related to the characteristics of the sample itself, and this phenomenon has been reported in the literatures before [22,23].

Download:
Fig. 3. (a) Adsorption isotherms of N2 at 77 K and CO2 at 195 K for SXU-121. (b) Adsorption isotherms for CO2 and N2 at 273 K and 298 K for SXU-121. (c) The isosteric heat of CO2 adsorption (Qst) for SXU-121 calculated according to the Clausius-Clapeyron equation. (d) Experimental column breakthrough curves for CO2/O2/N2 mixture (1/21/78, v/v/v) in an adsorber bed packed with SXU-121 at 298 K and 1 bar.

Although N2 adsorption cannot demonstrate the permanent porosity of SXU-121, the large CO2 adsorption capacity at ambient temperature does prove that SXU-121 possesses superior porosity (Fig. 3b). Since SXU-121 adsorbs almost no N2 even at 77 K, it is reasonable to assume that SXU-121 is highly selective for CO2 over N2 at ambient temperature. Consequently, we conducted a more detailed study on the single-component adsorption of CO2 and N2 at 273 K and 298 K (Fig. 3b). The maximum adsorption amounts of CO2 reach as high as 104.71 cm3/g and 63.56 cm3/g at 273 K and 298 K at 1 atm, respectively. This adsorption capacity is comparable to that of some well-known MOFs reported in the literature such as PCN-200, SIFSIX-3-Cu, SIFSXI-3-Zn (Table S2 in Supporting information) [24,25]. After the CO2 adsorption-desorption tests at 298 K, the framework structure of SXU-121 remains intact through its unchanged PXRD pattern (Fig. S3 in Supporting information). By contrast, the maximum uptake of N2 is only 12.13 cm3/g at 273 K. Specially, under the 298 K condition of our interest, SXU-121 reveals almost no adsorption of N2 (1.68 cm3/g), which is much less than CO2 as expected.

Ideal adsorbed solution theory (IAST) was applied to quantitatively calculate the selectivity of CO2 over N2. The calculated selectivity of CO2/N2 (15/85, v/v) mixtures at 298 K and 1 bar is as high as 9.51 × 107 (Fig. S4 in Supporting information). This value is quite high [23,2633], possibly due to the extremely low amount of N2 adsorption, which might cause some deviation in the calculated value (Table S2). However, this also indirectly indicates that the MOF has very high selectivity for CO2. The affinity toward guest molecules was evaluated by the heat of adsorption (Qst), according to Clausius-Clapeyron equation. On the basis of the adsorption isotherms at 273 K and 298 K (Fig. S7 in Supporting information), the calculated Qst of SXU-121 toward CO2 was slightly increased from 29.56 kJ/mol to 32.91 kJ/mol with CO2 uptake increasing from 0 to 2.8 mmol/g (Fig. 3c). Such level of Qst is moderate, indicating that the adsorption of CO2 is physisorption, which is crucial for reducing energy consumption during CO2 desorption in practical applications.

In order to validate the practical separation performance, breakthrough experiments have been performed and CO2/O2/N2 feed gas in a volume ratio of 1/21/78 flowed over a packed bed of SXU-121 with a flow rate of 2 mL/min at 298 K (Fig. 3d). The results demonstrated that SXU-121 can effectively separate CO2 from the mixture under ambient conditions. N2 and O2 were first eluted from the outlet with no detectable CO2 found. CO2 can be restored in the column until it broke at 9 min/g. From the above experiments, the designed SXU-121 can be deemed as a suitable material for one-step CO2/N2/O2 separation.

The capability of SXU-121 to adsorb single-component gases, specifically light hydrocarbons were also assessed at temperatures of 273 K (Fig. S5 in Supporting information) and 298 K (Fig. S6 in Supporting information). At 298 K and 1 atm, SXU-121 exhibited adsorption capacities of 64.45 cm3/g for C2H2, 58.63 cm3/g for C2H4, and 61.08 cm3/g for C2H6. SXU-121 did not demonstrate significant selective adsorption toward those gases. For C3H6 and C3H8, the adsorption capacities were 31.07 cm3/g and 8.74 cm3/g, respectively. Although both of the adsorption capacity are not high, they exhibit significant selective adsorption for C3H6 compared to C3H8. IAST calculation was also applied to estimate the selectivity of C3H6/C3H8 (50/50, v/v). The calculated selectivities of C3H6/C3H8 for equimolar binary mixtures at 298 K and 1 bar was 1.68 × 107 (Fig. S8 in Supporting information). Under identical conditions, the CH4 adsorption amount of SXU-121 is 18.20 cm3/g, which is much lower than that of C2H2, meaning that SXU-121 has a certain degree of selectivity to C2H2, and it has the prospect of biogas purification application. The calculated selectivity of C2H2/CH4 for equimolar binary mixtures at 298 K and 1 bar is 14.59 (Fig. S9 in Supporting information).

To elucidate the mechanism of interactions between SXU-121 framework and CO2 gas molecules, we employed Grand Canonical Monte Carlo (GCMC) simulations using the crystallographic information file of SXU-121 as the input model. The computational results, as shown in Fig. 4a, indicate that the adsorbed CO2 molecules were most preferentially adsorb between two open Cu(Ⅰ) sites. The distances between the two oxygen atoms of the CO2 molecule and the adjacent Cu(Ⅰ) are 2.356 Å and 2.436 Å, respectively. These distances are significantly smaller than the sum of the van der Waals radii of Cu and O (2.92 Å), indicating a very strong interaction between Cu(Ⅰ) sites and CO2 molecules. The distribution of adsorbed CO2 by SXU-121 were also simulated under various pressures (Figs. 4b-d). At low pressures (10 kPa), CO2 molecules preferentially adsorb around the Cu(Ⅰ) atoms, and as the pressure increases, the concentration of CO2 molecules within the 1D channels also increases. These results also indicate that CO2 is preferentially adsorbed near the Cu(Ⅰ) sites and once all the Cu(Ⅰ) sites are occupied, CO2 molecules are then adsorbed within the pores.

Download:
Fig. 4. The binding sites of CO2 for SXU-121. (a) The most preferential binding sites of adsorbed CO2 for SXU-121 at 298 K. (b-d) The simulated CO2 density distribution of SXU-121 at 10, 30, 90 kPa respectively at 298 K.

In conclusion, by employing one 1,3,5-triazine derivative (5-azacytosine) as organic linker, we have successfully constructed a Cu(Ⅰ) MOF SXU-121 with all organic linkers and metal ions three-connected to form a rare eta topology. SXU-121 features excellent water and thermal stability, as well as permanent porosity. SXU-121 exhibits rather high selectivity for CO2 at ambient temperature and achieves selective adsorption of CO2 under breakthrough experiment conditions over N2 and O2. Due to the diversity of 1,3,5-triazine derivatives and their ease of functionalization, we believe this will pave the way for synthesizing a class of metal-triazine framework materials with a wide range of functionalities.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Feng-Fan Yang: Writing – original draft, Validation, Methodology, Investigation, Data curation. Yin-Kang Ding: Investigation. Lin-Kai Wu: Methodology. Jiayue Tian: Validation, Software, Resources, Methodology. Shuai Dou: Software, Resources. Wenjing Wang: Validation, Software, Resources. Linfeng Liang: Writing – review & editing, Investigation, Funding acquisition.

Acknowledgment

This research was supported by National Natural Science Foundation of China (Nos. 22001154 and 22471147).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110550.

References
[1]
O. Delgado-Friedrichs, M. O'Keeffe, O.M. Yaghi, Acta Crystallogr. 62 (2006) 350-355.
[2]
M. Li, D. Li, M. O'Keeffe, O.M. Yaghi, Chem. Rev. 114 (2014) 1343-1370. DOI:10.1021/cr400392k
[3]
D. Zhao, D.J. Timmons, D.Q. Yuan, H.C. Zhou, Acc. Chem. Res. 44 (2011) 123-133. DOI:10.1021/ar100112y
[4]
B. Bann, S.A. Miller, Chem. Rev. 58 (1958) 131-172. DOI:10.1021/cr50019a004
[5]
M. Kalmutzki, M. Ströbele, F. Wackenhut, A.J. Meixner, H.J. Meyer, Angew. Chem. Int. Ed. 53 (2014) 14260-14263. DOI:10.1002/anie.201407708
[6]
Y. Kamakura, P. Chinapang, S. Masaoka, et al., J. Am. Chem. Soc. 142 (2020) 27-32. DOI:10.1021/jacs.9b10436
[7]
T.J. Koller, S.M.J. Endraß, M. Rösch, et al., Angew. Chem. Int. Ed. 63 (2024) e202404927.
[8]
S. Tominaka, H. Hamoudi, T. Suga, et al., Chem. Sci. 6 (2015) 1465-1473.
[9]
D. Sun, Y.H. Li, S.T. Wu, et al., CrystEngComm 13 (2011) 7311-7315. DOI:10.1039/c1ce05672g
[10]
H.B. Zhang, M.J. Zhang, P. Lin, et al., Chem. Eur. J. 22 (2016) 1141-1145. DOI:10.1002/chem.201503561
[11]
Y. Zhang, Q.F. Yang, X.P. Li, et al., CrystEngComm 22 (2020) 5690-5697. DOI:10.1039/d0ce00835d
[12]
M. Eddaoudi, D.F. Sava, J.F. Eubank, K. Adil, V. Guillerm, Chem. Soc. Rev. 44 (2015) 228-249.
[13]
K.C. Wang, Y.P. Li, L.H. Xie, X.Y. Li, J.R. Li, Chem. Soc. Rev. 51 (2022) 6417-6441. DOI:10.1039/d1cs00891a
[14]
J.P. Zhang, Y.B. Zhang, J.B. Lin, X.M. Chen, Chem. Rev. 112 (2012) 1001-1033. DOI:10.1021/cr200139g
[15]
X.W. Zhang, H. He, Y.W. Gan, et al., Angew. Chem. Int. Ed. 63 (2024) e202317648.
[16]
M.H. Almatarneh, A.A.A.A. Abu-Saleh, K.M. Uddin, R.A. Poirier, P.L. Warburton, Int. J. Quantum Chem. 117 (2017) 180-189. DOI:10.1002/qua.25308
[17]
M. Hatanaka, Spectrochim. Acta A 202 (2018) 87-92.
[18]
Y.H. Jang, S. Hwang, S.B. Chang, J. Ku, D.S. Chung, J. Phys. Chem. A 113 (2009) 13036-13040. DOI:10.1021/jp9053583
[19]
T.F. Willems, C.H. Rycroft, M. Kazi, J.C. Meza, M. Haranczyk, Microporous Mesoporous Mater. 149 (2012) 134-141.
[20]
A. Spek, J. Appl. Crystallogr. 36 (2003) 7-13.
[21]
P.S. Nugent, V.L. Rhodus, T. Pham, et al., J. Am. Chem. Soc. 135 (2013) 10950-10953. DOI:10.1021/ja4054948
[22]
J. Peng, Z. Liu, Y. Wu, S. Xian, Z. Li, ACS Appl. Mater. Interfaces 14 (2022) 21089-21097. DOI:10.1021/acsami.2c04779
[23]
Y. Shi, Y. Xie, H. Cui, et al., Chem. Engin. J. 446 (2022) 137101.
[24]
C. Xiao, J.D. Tian, Q.H. Chen, M.C. Hong, Chem. Sci. 15 (2024) 1570-1610. DOI:10.1039/d3sc06076d
[25]
D.D. Zhou, X.W. Zhang, Z.W. Mo, et al., EnergyChem 1 (2019) 100016.
[26]
H. Alawisi, B. Li, Y. He, et al., Cryst. Growth Des. 14 (2014) 2522-2526. DOI:10.1021/cg500235j
[27]
Z.L. Ma, P.X. Liu, Z.Y. Liu, et al., Inorg. Chem. 60 (2021) 6550-6558. DOI:10.1021/acs.inorgchem.1c00357
[28]
P. Nugent, Y. Belmabkhout, S.D. Burd, et al., Nature 495 (2013) 80-84. DOI:10.1038/nature11893
[29]
J. Peng, Z. Liu, Y. Wu, et al., ACS Appl. Mater. Interfaces 14 (2022) 21089-21097. DOI:10.1021/acsami.2c04779
[30]
B. Wang, H. Huang, X.L. Lv, et al., Inorg. Chem. 53 (2014) 9254-9259. DOI:10.1021/ic5013473
[31]
Z.S. Wang, M. Li, Y.L. Peng, et al., Angew. Chem. Int. Ed. 58 (2019) 16071-16076. DOI:10.1002/anie.201909046
[32]
F. Yang, X. Wang, J. Tian, X. Wang, L. Liang, Processes 11 (2023) 1387. DOI:10.3390/pr11051387
[33]
B. Zheng, H. Liu, Z. Wang, et al., CrystEngComm 15 (2013) 3517-3520. DOI:10.1039/c3ce26177h