Water Based Synthesis of ZIF-8 Assisted by Hydrogen Bond Acceptors and Enhancement of CO2 Uptake by Solvent Assisted Ligand Exchange

The aim of this work was to synthesize zeolitic imidazolate framework-8 (ZIF-8) by an alternative method and then modify the surface properties for enhancing the CO2 adsorption performance. The ZIF-8 was synthesized by a water based synthesis method using 2-methyl imidazole (2-MeIM) as a hydrogen bond donor and quaternary ammonium salts (QAS) as a hydrogen bond acceptor. The optimal synthesis conditions were investigated by varying (i) the order of precursor mixing during the synthesis process (ii) different QAS (tetrabutyl ammonium bromide (TBAB), tetraethyl ammonium bromide (TEAB) and trimethyl phenyl ammonium bromide (TMPAB)) and (iii) the ratio between 2-MeIM and QAS. The results show that the optimal synthesis condition was using TMPAB as the hydrogen bond acceptor with the ratio between 2-MeIM and TMPAB of 8:2 and in the order of first mixing both hydrogen bond donor and acceptor before adding Zn(NO3)2·6H2O solution. TMPAB can provide uniform size distribution with the smallest particle sizes of ZIF-8. This can be explained by the higher hydrogen bond strength between hydrogen bond donor (2-MeIM) and hydrogen bond acceptor (TMPAB) when compared with that of the rest of two QAS. The synthesized ZIF-8 was modified by solvent-assisted ligand exchange methods. The organic linker of ZIF-8 (2-MeIM) was exchanged by 2-aminobenzimidazole (2-NH2bZIM) and 2-phenylimidazole (2-PhIM). The CO2 uptake of modified ZIF-8 was enhanced upon exchanging with 2-NH2bZIM. The increase in CO2 uptake was due to an additional interaction between CO2 and exchanged imidazole linker and an increase in surface properties (higher surface area, pore size and pore volume).


Introduction
In recent years, global warming and climate change have become the most important problems facing humanity. These phenomena are caused by large amounts of greenhouse gases emission into the atmosphere. Carbon dioxide (CO 2 ) is one of the greenhouse gases which is mainly generated from burning fossil fuels in industrial processes for the generation of energy and transportation. Therefore, a decrease in CO 2 emissions is needed, which can be achieved by reducing fossil fuel burning and by capturing CO 2 from the atmosphere. Various technologies are used for capturing CO 2 , such as cryogenic separation [1][2][3], CO 2 membrane-based separation [4][5][6], chemical absorption [7][8][9] and adsorption technology [10][11][12]. Solid adsorption technology is now of interest due to its many advantages, such as energy-efficiency, low prices and the ease by which it can be modified. Several solid adsorbents which are porous have been studied for CO 2 adsorption, such as zeolites [13][14][15]

ZIF-8 Synthesis by Water Based Synthesis Assisted by Hydrogen Bond Acceptors
The molar ratio of precursor between Zn(NO 3 ) 2 ·6H 2 O and 2-MeIM was kept constant at 1:4. The effect of synthesis conditions on ZIF-8 properties were investigated. First, the effects of changing the ordering steps of precursor mixing were studied by using two different routes. For route no.1, Zn(NO 3 ) 2 ·6H 2 O (0.86 g, 0.00289 mol), 2-MeIM (0.95 g, 0.0116 mol) and TMPAB (0.62 g, 0.0287 mol) (TMPAB was selected as representative QAS sources) were mixed and heated under stirring for 30 min or until the precursor turned into a homogeneous solution. Then, the deionized water (50 mL, 2.8 mol) was added and the mixture was continuously stirred for 30 min; in this step, the solution turned turbid. After that, the mixture was centrifuged at 6000 rpm for 10 min, after which the white precipitates were obtained. The precipitant was washed with deionized water several times. Finally, the solid sample was dried at 120 • C for 24 hours. For route no.2, 2-MeIM and QAS were first mixed and heated under continuous stirring for 30 min or until a homogeneous solution was obtained. Zn(NO 3 ) 2 ·6H 2 O which was dissolved in 50 mL of deionized water was added into the former solution and stirred for 30 min. After that, the synthesis procedure was carried out by the same process as route no.1. Second, the effects of different types of QAS were studied by using three different sources (tetrabutyl ammonium bromide (TBAB), tetraethyl ammonium bromide (TEAB) and trimethyl phenyl ammonium bromide (TMPAB)). The ratio between 2-MeIm and QAS was kept constant at 8:2. In synthesis procedure, 2-MeIM and QAS were first mixed together to form a homogeneous solution before the addition of the zinc nitrate solution. The temperature at homogeneous point was defined as synthesis temperature. The synthesis temperature of 2-MeIM and TBAB, TEAB and TMPAB were 109.0, 80.0 and 120.0 • C, respectively. Third, the effects of various ratios between 2-MeIM and QAS were studied by varying the ratios of these two precursors, at 4:6, 5:5, 6:4, 7:3 and 8:2.
Moreover, ZIF-8 was also synthesized by the solvothermal method for comparison with synthesized ZIF-8 by the above method. The molar ratio between Zn precursor and 2-MeIM was 1:4, and methanol was used as a solvent. First, Zn(NO 3 ) 2 ·6H 2 O and 2-MeIM were separately dissolved in methanol and stirred until the precursors were completely dissolved. Second, two solutions were mixed and continuously stirred for 1 hour. The mixture was centrifuged at 6000 rpm for 20 min to obtain the white precipitates. The solid was filtered and washed with deionized water and methanol, respectively. Finally, the solid sample was dried at 80 • C for 24 hours.

Modification of Synthesized ZIF-8 by Solvent-assisted Ligand Exchange (SALE) Method
In this modification method, 2-MeIM linker in synthesized ZIF-8 was exchanged with two imidazole derivatives (2-NH 2 bZIM and 2-PhIM). The preparation details followed C. W. Tsai et al. [37]. Briefly, 100 mg of synthesized ZIF-8 was dispersed in methanol for 3 min under ultrasonic probe. Then the solution was transferred into a polypropylene bottle. At the same time, the imidazole derivatives were also dissolved in appropriate solvents (methanol and n-butanol for 2-NH 2 bZIM and 2-PhIM, respectively). The imidazole derivative solution was then added into the ZIF-8 solution and the mixture was heated at 60 • C for 3 days to exchange the linker. After cooling the solution, the obtained solid was washed by methanol and then separated by centrifugation at 6000 rpm for 15 min. Finally, the modified ZIF-8 was obtained by drying at 60 • C for 24 hours. The effect of molar ratio between synthesized ZIF-8 and imidazole derivatives were varied at 1:1, 1:3 and 1:5.

Characterization
To confirm and identify the structures of synthesized and modified samples, powder X-ray diffraction (PXRD) patterns were obtained by a PANalytical Empyrean X-ray diffractometer (Malvern Panalytical, Malvern, UK). The 2θ was operated in the range of 5 • to 50 • .
The morphology of synthesized and modified ZIF-8 was determined by a LEO 1450VP scanning electron microscope (SEM) (Carl Zeiss AG, Oberkochen, Germany). The powder sample was prepared before scanning by dispersing it on carbon tape that was put on a specimen stub. After that, the sample was coated with a thin layer of gold by low-vacuum sputter coating to reduce the electron charge.
The surface and porosity properties (surface area, pore volume and average pore size) of synthesized and modified samples were determined by measuring the nitrogen adsorption/desorption isotherm at 77 K by Bel Sorp mini II (MicrotracBEL Corp., Osaka, Japan) The sample was degassed at 120 • C for 5 hours before the adsorption/desorption process to remove water molecules or impure molecules from the sample surface. The BET equation was used to calculate the specific surface area at a relative pressure (P/P 0 ) range between 0 and 0.1. The pore volume of samples was determined at the relative pressure close to unity (P/P 0 ≈ 1).
To evidence the exchange achieved during the SALE process, the modified samples were analyzed by digestive liquid 1 H-NMR in which acetic acid-d4 was used as a solvent. 1 H-NMR spectra were obtained by Bruker AVANCE III 500 MHz (Bruker Corp., Billerica, MA, USA).

CO 2 Adsorption Study
The adsorption measurement was performed using Pyris Diamond TG/DTA Perkin Elmer Analysis. The CO 2 uptake was measured at 40 • C and ambient pressure. The surface of synthesized ZIF-8 was pre-treated by heating (heating rate 10 • C/min) the sample at 150 • C under purging of N 2 for 45 min (flow rate 100 mL/min). After the pretreatment process, the sample was cooled down to 40 • C under N 2 atmosphere. After that, pure CO 2 was then switched to the sample to start the gas uptake process and the CO 2 uptake was proceeded for 95 min. Upon decreasing the temperature, the weight of the sample increased due to the adsorption of CO 2 on the solid surface. The CO 2 uptake can be calculated by differentiating the weight before and after adsorption.

Effect of Mixed Ordering
Step on ZIF-8 Properties All precursors were mixed together under heating and stirring to form a final product. However, the order of precursor mixing also dominated the obtained samples. Therefore, the effect of the order of mixing between the precursors on ZIF-8 properties was investigated by using two different routes; route 1 and 2. For route 1, all precursors (Zn(NO 3 ) 2 ·6H 2 O, 2-MeIM and TMPAB) were mixed together under heating and stirring. After that, deionized water was subsequently added to precipitate the product. For route 2, 2-MeIM and QAS were firstly mixed and the Zn(NO 3 ) 2 ·6H 2 O solution was then added. Figure 1 illustrates the XRD patterns of synthesized ZIF-8 obtained from routes 1 and 2 compared with that from solvothermal method. It is clearly seen that only the synthesized sample Crystals 2020, 10, 599 5 of 23 by route 2 can provide the characteristic peaks of ZIF-8 which is in accordance with ZIF-8 by the solvothermal method. However, ZIF-8 cannot be formed by route 1. Consequently, the optimum method for the mixing sequence of precursors was route 2. Moreover, the role of QAS was also investigated by synthesizing ZIF-8 without the addition of QAS. The results show that ZIF-8 cannot be produced without adding QAS by this synthesis procedure, as illustrated in XRD patterns in Figure 1 (2-MeIM: Zn 2+ sample). This indicated that QAS acts as a template for ZIF-8 formation during the synthesis process [31].
Crystals 2020, 10, x FOR PEER REVIEW 5 of 23 method for the mixing sequence of precursors was route 2. Moreover, the role of QAS was also investigated by synthesizing ZIF-8 without the addition of QAS. The results show that ZIF-8 cannot be produced without adding QAS by this synthesis procedure, as illustrated in XRD patterns in Figure 1 (2-MeIM: Zn 2+ sample). This indicated that QAS acts as a template for ZIF-8 formation during the synthesis process [31].

Effect of Different Quaternary Ammonium Salts
From the previous part, the optimal mixing order of precursors was route 2, i.e., the organic linker and QAS were first mixed and Zn(NO3)2⋅6H2O solution was then added. In this part, the effects of different quaternary ammonium salts on ZIF-8 properties were investigated. Figure 2 displays XRD patterns of synthesized samples obtained from three different QAS: tetrabutyl ammonium bromide, tetraethyl ammonium bromide and trimethyl phenyl ammonium bromide. At the same ratio between 2-MeIM and QAS (8:2), tetraethyl ammonium bromide, tetrabutyl ammonium bromide and trimethyl phenyl ammonium bromide can provide yielded ZIF-8 with different crystallinities. Synthesized ZIF-8 obtained from TMPAB gave the highest characteristic peak intensities of ZIF-8, which can attribute to high crystallinity. The SEM results confirmed the phase, crystal shape and morphologies of all synthesized samples. Figure 3 displays SEM images of ZIF-8 obtained from three different QAS. ZIF-8 obtained from tetraethyl ammonium bromide and tetrabutyl ammonium bromide exhibited an irregular shape with a high degree of agglomeration. While ZIF-8 obtained from trimethyl phenyl ammonium bromide showed a well distributed particle of truncated rhombic dodecahedra shape with a significant reduction in particle size lower than that of ZIF-8 obtained from both former QAS. Therefore, TMPAB was selected to be used as the QAS (hydrogen bond acceptor) for the further synthesis process in the next part.

Effect of Different Quaternary Ammonium Salts
From the previous part, the optimal mixing order of precursors was route 2, i.e., the organic linker and QAS were first mixed and Zn(NO 3 ) 2 ·6H 2 O solution was then added. In this part, the effects of different quaternary ammonium salts on ZIF-8 properties were investigated. Figure 2 displays XRD patterns of synthesized samples obtained from three different QAS: tetrabutyl ammonium bromide, tetraethyl ammonium bromide and trimethyl phenyl ammonium bromide. At the same ratio between 2-MeIM and QAS (8:2), tetraethyl ammonium bromide, tetrabutyl ammonium bromide and trimethyl phenyl ammonium bromide yielded ZIF-8 with different crystallinities. Synthesized ZIF-8 obtained from TMPAB gave the highest characteristic peak intensities of ZIF-8, which can attribute to high crystallinity. The SEM results confirmed the phase, crystal shape and morphologies of all synthesized samples. Figure 3 displays SEM images of ZIF-8 obtained from three different QAS. ZIF-8 obtained from tetraethyl ammonium bromide and tetrabutyl ammonium bromide exhibited an irregular shape with a high degree of agglomeration. While ZIF-8 obtained from trimethyl phenyl ammonium bromide showed a well distributed particle of truncated rhombic dodecahedra shape with a significant reduction in particle size lower than that of ZIF-8 obtained from both former QAS. Therefore, TMPAB was selected to be used as the QAS (hydrogen bond acceptor) for the further synthesis process in the next part. Crystals 2020, 10, x FOR PEER REVIEW 6 of 23  The role of quaternary ammonium salts in synthesis process was to accelerate the deprotonation of 2-MeIM and the deprotonated 2-MeIM was further coordinated with the metal ion (Zn 2+ ) and forming the framework [31]. This effect can accelerate the nucleation and crystal growth rate. Finally, ZIF-8 with a uniform distribution of particle sizes was produced. The deprotonation can be accelerated by facilitating withdrawing the hydrogen atom from 2-MeIM, which can be done by forming a hydrogen bond with a hydrogen bond acceptor. In this work, 2-methyle imidazole acted as a hydrogen bond donor, while quaternary ammonium salts played a role as a hydrogen bond acceptor. Three quaternary ammonium salts had different hydrogen bond acceptor properties due to their different structures. Therefore, hydrogen bonding between 2-MeIM and QAS was also different. The hydrogen bond strength between these two precursors can dominate the deprotonation of the hydrogen bond donor (2-MeIM) and thus the nucleation and crystal growth of ZIF-8 can be controlled.
Hydrogen bonding is an attractive interaction between hydrogen bond donor (denoted as HX) and hydrogen bond acceptor (denoted as Y); thus, hydrogen bonding can be expressed as XH···Y, where X is an electronegative atom, such as O, N or F, and Y can be any electron rich atom. C. Laurence and M. Berthelot [38] studied the strength of hydrogen bond acceptors (HBAs) by measuring from the Gibbs energy change for the formation of 1:1 hydrogen bonding complexes between hydrogen bond acceptors (bases) and a reference hydrogen bond donor (4-fluorophenol) in CCl 4 at 298 K. They concluded that the HBA strength depended on three factors: (i) the acceptor atom position in the periodic table; (ii) the polarizability and resonance effect of substituents around the acceptor atom; and (iii) the steric hindrance of the acceptor site.

281
The role of quaternary ammonium salts in synthesis process was to accelerate the deprotonation  Figure 4 illustrates the structure of three quaternary ammonium salts. The hydrogen bonding between 2-MeIM and these three QAS occurred at the bromine ion (Br − ) site. Therefore, Br − ion is an acceptor site. Since three ammonium bromide salts had a different substituents group around Br − ion, their hydrogen bond strength to 2-MeIM must be also different. Because the hydrogen bond donor was 2-MeIM and the hydrogen bond acceptors were varied (TEAB, TBAB and TMPAB), the hydrogen bond strength depended on the hydrogen bond acceptor strength (HBAs). acceptor atom; and iii) the steric hindrance of the acceptor site. Figure 4 illustrates the structure of three quaternary ammonium salts. The hydrogen bonding between 2-MeIM and these three QAS occurred at the bromine ion (Br − ) site. Therefore, Brion is an acceptor site. Since three ammonium bromide salts had a different substituents group around Brion, their hydrogen bond strength to 2-MeIM must be also different. Because the hydrogen bond donor was 2-MeIM and the hydrogen bond acceptors were varied (TEAB, TBAB and TMPAB), the hydrogen bond strength depended on the hydrogen bond acceptor strength (HBAs).  The results from Figures 2 and 3 show that ZIF-8 obtained by using TMPAB exhibited the most uniform size distribution with the smallest particle size when compared with the other two QAS. This can be explained by the different substituent groups around the acceptor atom which dominated the hydrogen bond strength. Since the hydrogen bonding of 2-MeIM with three QAS occurred at the same acceptor site (Br − ion), the substituent groups around bromine ions must dominate the hydrogen bond strength. The substituents functional groups around bromine ions was the ammonium group. This ammonium group also had different substituents functional groups, i.e., alkane for both TEAB and TBAB and methyl with phenyl groups for TMPAB. Considering the steric effect, TMPAB had the lowest steric hindrance around the Brion when compared with TEAB and TBAB. Therefore, TMPAB should exhibit the highest hydrogen bond strength with 2-MeIM. Besides this interaction at the Br − ion acceptor site, an additional interaction between TMPAB and 2-MeIM was also reported, i.e., The results from Figures 2 and 3 show that ZIF-8 obtained by using TMPAB exhibited the most uniform size distribution with the smallest particle size when compared with the other two QAS. This can be explained by the different substituent groups around the acceptor atom which dominated the hydrogen bond strength. Since the hydrogen bonding of 2-MeIM with three QAS occurred at the same acceptor site (Br − ion), the substituent groups around bromine ions must dominate the hydrogen bond strength. The substituents functional groups around bromine ions was the ammonium group. This ammonium group also had different substituents functional groups, i.e., alkane for both TEAB and TBAB and methyl with phenyl groups for TMPAB. Considering the steric effect, TMPAB had the lowest steric hindrance around the Br − ion when compared with TEAB and TBAB. Therefore, TMPAB should exhibit the highest hydrogen bond strength with 2-MeIM. Besides this interaction at the Br − ion acceptor site, an additional interaction between TMPAB and 2-MeIM was also reported, i.e., interaction between phenyl group and 2-MeIM. M. Levitt and M. F. Perutz [39] reported that aromatic rings can act as hydrogen bond acceptors. They revealed that there is a significant interaction between a hydrogen bond donor (such as NH group) and the center (π-electron) of a benzene ring, which acts as a hydrogen bond acceptor. Therefore, the phenyl group in TMPAB can also interact with 2-MeIM by hydrogen bonding as mentioned above. Consequently, hydrogen bonding between 2-MeIM and TMPAB can occur at both Br − ion and phenyl acceptor sites, as shown in Figure 5. From these reasons, using TMPAB to promote the deprotonation of 2-MeIM can provide the ZIF-8 with high uniform size distribution and the smallest ZIF-8 particle sizes. Figure 5 illustrates the proposed ZIF-8 formation mechanism with the use of trimethyl phenyl ammonium bromide (TMPAB) as a hydrogen bond acceptor. The hydrogen bonding between 2-MeIM can occur at two acceptor sites; the Br − ion and π-electron sites. From this effect, the hydrogen bond strength was increased and the deprotonation of 2-MeIM was promoted. Therefore, the formation of ZIF-8 framework by coordination of deprotonated 2-MeIM and Zn 2+ ion was also accelerated and finally the well size distribution of ZIF-8 particle was controlled and obtained. These results indicate that the additional hydrogen bonding between the π-electron site of benzene ring in TMPAB and 2-MeIM played an important role in enhancing the hydrogen bond strength and resulted in increasing the deprotonation of 2-MeIM process. On the other hand, from the results of XRD and SEM, ZIF-8 obtained from TBAB exhibited higher crystallinity than that of TEAB. This due to the TBAB structure which is more sterically hindered than TEAB and can lead to less association between the negative Br − ion and the positive N atom in its structure; therefore, the Br − ion was more available for bonding with 2-MeIM. 8 particle was controlled and obtained. These results indicate that the additional hydrogen bonding between the π-electron site of benzene ring in TMPAB and 2-MeIM played an important role in enhancing the hydrogen bond strength and resulted in increasing the deprotonation of 2-MeIM process. On the other hand, from the results of XRD and SEM, ZIF-8 obtained from TBAB exhibited higher crystallinity than that of TEAB. This due to the TBAB structure which is more sterically hindered than TEAB and can lead to less association between the negative Brion and the positive N atom in its structure; therefore, the Brion was more available for bonding with 2-MeIM.  Figure 5. Proposed ZIF-8 formation mechanism with the use of trimethylphenyl ammonium bromide (TMPAB) as hydrogen bond acceptor.

Effect of Ratio between 2-MeIM and QAS
In this part, the ratio between 2-MeIM and TMPAB was varied at 5:5, 6:4, 7:3 and 8:2. It is noteworthy that the ratio between these two precursors at 4:6 cannot provide the desired product. From Figure 6, it is seen that all ratios exhibited ZIF-8 characteristic peaks with different intensities. Figure 7 shows SEM images of synthesized ZIF-8 for all ratios. The particle size of synthesized ZIF-8 was reduced upon lowering the QAS content. The particle size distribution of all samples is also displayed in the inset of each image. The size distributions were in the range of 500-1500 nm for 5:5, Figure 5. Proposed ZIF-8 formation mechanism with the use of trimethylphenyl ammonium bromide (TMPAB) as hydrogen bond acceptor.

Effect of Ratio between 2-MeIM and QAS
In this part, the ratio between 2-MeIM and TMPAB was varied at 5:5, 6:4, 7:3 and 8:2. It is noteworthy that the ratio between these two precursors at 4:6 cannot provide the desired product. From Figure 6, it is seen that all ratios exhibited ZIF-8 characteristic peaks with different intensities. Figure 7 shows SEM images of synthesized ZIF-8 for all ratios. The particle size of synthesized ZIF-8 was reduced upon lowering the QAS content. The particle size distribution of all samples is also displayed in the inset of each image. The size distributions were in the range of 500-1500 nm for 5:5, 6:4 and 7:3, while uniform size distribution of particle sizes of 400-600 nm was found for the sample with the ratio of 8:2.
In order to confirm the bonding in synthesized ZIF-8, ZIF-8 obtained from an optimal synthesis condition was used as a representative for all ZIF-8 samples. Figure 8 illustrates FTIR of original ZIF-8 synthesized by water based synthesis assisted by hydrogen bond acceptors (the ratio between 2-MeIM and TMPAB as 8:2). It was found that the infrared (IR) spectra of the ZIF-8 sample was in a good agreement with other works [40,41]. The peak at 423 cm −1 was assigned as Zn-N stretching vibration. The 690 cm −1 peak was attributed to the ring out of plane bending vibration of 2-MeIM. The 998 and 762 cm −1 peaks were due to C-N bending vibration and C-H bending mode. The peak at 1140 cm −1 was due to an aromatic C-N stretching mode, while the peak at 1607 cm −1 resulted from C=N stretching vibration. The peaks around 1305 to 1436 cm −1 were assigned as entire ring stretching, and the peak at 1513 cm −1 was due to the C=C stretching mode. The peaks at 2917 and 3100 cm −1 were attributed as aliphatic and aromatic C-H stretching of methyl imidazole, respectively.
vibration. The 690 cm −1 peak was attributed to the ring out of plane bending vibration of 2-MeIM. The 998 and 762 cm −1 peaks were due to C-N bending vibration and C-H bending mode. The peak at 1140 cm −1 was due to an aromatic C-N stretching mode, while the peak at 1607 cm −1 resulted from C=N stretching vibration. The peaks around 1305 to 1436 cm −1 were assigned as entire ring stretching, and the peak at 1513 cm −1 was due to the C=C stretching mode. The peaks at 2917 and 3100 cm −1 were attributed as aliphatic and aromatic C-H stretching of methyl imidazole, respectively.

XRD and SEM
Synthesized ZIF-8 obtained from TMPAB with the ratio of 8:2 was selected to be used for modification by solvent-assisted ligand exchange method. The 2-methyl imidazole linker was replaced by two imidazole derivatives (2-NH 2 bZIM and 2-PhIM). The molar ratio between synthesized ZIF-8 and imidazole derivatives was varied at 1:1, 1:3 and 1:5. Figure 9a,b illustrate XRD patterns of modified ZIF-8 by exchanging with 2-NH 2 bZIM and 2-PhIM, respectively. A modified ZIF-8 with both imidazole derivatives exhibited the ZIF-8 characteristic peaks, which indicated that the ZIF-8 framework was preserved upon replacing 2-MeIM with two imidazole derivatives. Upon increasing the imidazole derivatives, the ZIF-8 structure remained unchanged while the peak intensities were different, which indicated different crystallinities of ZIF-8 crystals. Figure 10 displays SEM images of exchanged ZIF-8 with 2-NH 2 bZIM ( Figure 10a) and 2-PhIM ( Figure 10b) at various ratios between ZIF-8 and two imidazole derivatives. All samples exhibited larger particle sizes upon exchanging with two imidazole derivatives, i.e., the particle size of original ZIF-8 was around 400-600 nm, while those of all exchanged samples were in the range of 600-1500 nm. A few changes in the particle shapes were observed which was due to agglomeration.

XRD and SEM
Synthesized ZIF-8 obtained from TMPAB with the ratio of 8:2 was selected to be used for modification by solvent-assisted ligand exchange method. The 2-methyl imidazole linker was replaced by two imidazole derivatives (2-NH2bZIM and 2-PhIM). The molar ratio between synthesized ZIF-8 and imidazole derivatives was varied at 1:1, 1:3 and 1:5. Figure 9a, b illustrate XRD patterns of modified ZIF-8 by exchanging with 2-NH2bZIM and 2-PhIM, respectively. A modified ZIF-8 with both imidazole derivatives exhibited the ZIF-8 characteristic peaks, which indicated that the ZIF-8 framework was preserved upon replacing 2-MeIM with two imidazole derivatives. Upon increasing the imidazole derivatives, the ZIF-8 structure remained unchanged while the peak intensities were different, which indicated different crystallinities of ZIF-8 crystals.

XRD and SEM
Synthesized ZIF-8 obtained from TMPAB with the ratio of 8:2 was selected to be used for modification by solvent-assisted ligand exchange method. The 2-methyl imidazole linker was replaced by two imidazole derivatives (2-NH2bZIM and 2-PhIM). The molar ratio between synthesized ZIF-8 and imidazole derivatives was varied at 1:1, 1:3 and 1:5. Figure 9a, b illustrate XRD patterns of modified ZIF-8 by exchanging with 2-NH2bZIM and 2-PhIM, respectively. A modified ZIF-8 with both imidazole derivatives exhibited the ZIF-8 characteristic peaks, which indicated that the ZIF-8 framework was preserved upon replacing 2-MeIM with two imidazole derivatives. Upon increasing the imidazole derivatives, the ZIF-8 structure remained unchanged while the peak intensities were different, which indicated different crystallinities of ZIF-8 crystals.

1 H NMR Analysis for the Replacement of Organic Linkers
In this section, ZIF-8 sample with the ratio of 1:3 was studied to confirm the replacement of 2-MeIM linker in ZIF-8 by 2-aminobenzimidazole and 2-phenylimidazole by the analysis of 1 H NMR spectral data (Figures 11-13). The signals at around 2.15 and 11.52 ppm which appeared in all samples were CH 3 and OH protons of digestive solvent, acetic acid-d 4 . In Figure 11, spectral 11A demonstrated the ZIF-8 which clearly showed the key combination signals of N-CH 3 and sets of aromatic protons (3.40, and 7.58, 7.65 and 7.75 ppm) of TMPAB salt and CH 3 and CH=CH (2.55 and 7.50 ppm) of 2-MeIM, which corresponded to spectral 11B and 11C. It is worthy to note that the TMPAB salt signals in all samples were shifted compared to the pure TMPAB salt (3.90, 7.68, 7.74 and 8.16 ppm, spectral 11B). This is due to the environmental changes that caused the shielding of TMPAB proton signals. For the 2-aminobenzimidazole exchanged sample (Figure 12, spectral 12A), the characteristic proton signals of 2-MeIM were clearly observed (2.55 and 7.50 ppm). Additionally, it showed phenyl proton signals of 2-NH 2 bZIM (7.31 and 7.49 ppm), which are comparable to spectral 12C. This indicated that 2-MeIM was replaced by 2-NH2bZIM during the SALE process. The relative ratio of ZIF-8 and exchanged ZIF-8 by 2-NH 2 bZIM was 1:0.11, based on the integration ratio of CH=CH protons of 2-MeIM (7.50 ppm, 2H) and the two ortho phenyl protons (7.31 ppm, 2H) of 2-NH 2 bZIM. It is also revealed in spectral 12A that the TMPAB salt signal intensity was decreased when compared to spectral 12B (7.65 and 7.75 ppm). The TMPAB salt may be removed during washing step of preparation process. For 2-phenylimidazole ( Figure 13, spectral 13A), again the signals of 2-MeIM were observed (2.55 and 7.50 ppm). The key signals of ortho-substituted protons (2 ,6 ) and imidazole aromatic protons (4,5) were observed at 8.10 and 7.72 ppm, which confirmed that 2-MeIM was replaced by 2-PhIM in SALE process. The approximately exchanged ratio was 1:0.04, based on the integration ratio of CH=CH protons of 2-MeIM (7.50 ppm, 2H) and the ortho-substituted protons (8.10 ppm, 2H) of 2-PhIM. In addition, the TMPAB salt signals were decreased when compared with spectral 13B (3.40 and 7.65 ppm).
Crystals 2020, 10, x FOR PEER REVIEW 15 of 23 For the 2-aminobenzimidazole exchanged sample (Figure 12, spectral 12A), the characteristic proton signals of 2-MeIM were clearly observed (2.55 and 7.50 ppm). Additionally, it showed phenyl proton signals of 2-NH2bZIM (7.31 and 7.49 ppm), which are comparable to spectral 12C. This indicated that 2-MeIM was replaced by 2-NH2bZIM during the SALE process. The relative ratio of ZIF-8 and exchanged ZIF-8 by 2-NH2bZIM was 1:0.11, based on the integration ratio of CH=CH protons of 2-MeIM (7.50 ppm, 2H) and the two ortho phenyl protons (7.31 ppm, 2H) of 2-NH2bZIM. It is also revealed in spectral 12A that the TMPAB salt signal intensity was decreased when compared to spectral 12B (7.65 and 7.75 ppm). The TMPAB salt may be removed during washing step of preparation process. For 2-phenylimidazole ( Figure 13, spectral 13A), again the signals of 2-MeIM were observed (2.55 and 7.50 ppm). The key signals of ortho-substituted protons (2',6') and imidazole aromatic protons (4,5) were observed at 8.10 and 7.72 ppm, which confirmed that 2-MeIM was replaced by 2-PhIM in SALE process. The approximately exchanged ratio was 1:0.04, based on the integration ratio of CH=CH protons of 2-MeIM (7.50 ppm, 2H) and the ortho-substituted protons (8.10 ppm, 2H) of 2-PhIM. In addition, the TMPAB salt signals were also decreased when compared with spectral 13B (3.40 and 7.65 ppm).

CO2 Uptake of ZIF-8 Synthesized from Water based Synthesis Assisted by Hydrogen Bond Acceptors
The CO2 uptake of all synthesized samples was measured by thermogravimetric analysis at 313 K and ambient pressure which were plotted as a function of time, as illustrated in Figure 14. The

CO 2 Uptake of ZIF-8 Synthesized from Water based Synthesis Assisted by Hydrogen Bond Acceptors
The CO 2 uptake of all synthesized samples was measured by thermogravimetric analysis at 313 K and ambient pressure which were plotted as a function of time, as illustrated in Figure 14.
The synthesized ZIF-8 with different ratios between 2-MeIM and TMPAB exhibited slightly different degree of CO 2 uptake abilities. The synthesized ZIF-8 sample with a ratio of 8:2 exhibited the highest CO 2 uptake when compared with other samples and the uptake amount was around 0.3 mmol/g at the adsorption time of 95 min. The CO 2 uptake of all samples was in the order of ZIF-8 (8:2) > ZIF-8 (7:3) > ZIF-8 (6:4) > ZIF-8 (5:5), which corresponded with the results of SEM, in that the particle size increased with a decrease in QAS content, i.e., the lower CO 2 uptake, the larger ZIF-8 particle sizes. The CO 2 uptake of ZIF-8 obtained by solvothermal method was also compared in Figure 14. It can be seen that the CO 2 uptake of ZIF-8 from the solvothermal method exhibited much higher CO 2 uptake than ZIF-8 synthesized from TMPAB (the water based synthesis method). The CO 2 adsorption behavior and CO 2 adsorption mechanism will be discussed in the next part.
increased with a decrease in QAS content, i.e., the lower CO2 uptake, the larger ZIF-8 particle sizes. The CO2 uptake of ZIF-8 obtained by solvothermal method was also compared in Figure 14. It can be seen that the CO2 uptake of ZIF-8 from the solvothermal method exhibited much higher CO2 uptake than ZIF-8 synthesized from TMPAB (the water based synthesis method). The CO2 adsorption behavior and CO2 adsorption mechanism will be discussed in the next part. 3.3.2. CO2 Uptake of Exchanged ZIF-8 by Imidazole Derivatives Figure 15a, b illustrate CO2 uptake of modified ZIF-8 by exchanging with 2-aminobenzimidazole and 2-phenylimidazole, respectively. The 2-NH2bZIM and 2-PhIM exchanged samples with different ratios between ZIF-8 and these two imidazole derivatives were also compared with the original synthesized ZIF-8. Upon exchanging the 2-MeIM of ZIF-8 by 2-NH2bZIM and 2-PhIM, the CO2 uptake was improved. The results also revealed that the increasing ratio of both imidazole derivatives have a slightly impact on the CO2 uptake of samples. Therefore, the ZIF-8 sample with the ratio of 1:3 was selected for further discussion of the CO2 adsorption behavior in the exchanged sample. 3.3.2. CO 2 Uptake of Exchanged ZIF-8 by Imidazole Derivatives Figure 15a,b illustrate CO 2 uptake of modified ZIF-8 by exchanging with 2-aminobenzimidazole and 2-phenylimidazole, respectively. The 2-NH 2 bZIM and 2-PhIM exchanged samples with different ratios between ZIF-8 and these two imidazole derivatives were also compared with the original synthesized ZIF-8. Upon exchanging the 2-MeIM of ZIF-8 by 2-NH 2 bZIM and 2-PhIM, the CO 2 uptake was improved. The results also revealed that the increasing ratio of both imidazole derivatives have a slightly impact on the CO 2 uptake of samples. Therefore, the ZIF-8 sample with the ratio of 1:3 was selected for further discussion of the CO 2 adsorption behavior in the exchanged sample. Figure 16 shows the CO 2 uptake of ZIF-8 and two ZIF-8 exchanged samples obtained from the water based synthesis method, and ZIF-8 obtained from the solvothermal method. It is clearly seen that upon exchanging the 2-MeIM linker with imidazole derivatives, the CO 2 uptake was enhanced. In order to explain the CO 2 uptake ability of synthesized ZIF-8 by hydrogen bond acceptor assisted and linker exchanged samples, data analysis of the surface properties for those samples were performed. Figure 17 shows N 2 adsorption desorption of original ZIF-8 (with the ratio of 8:2) and exchanged ZIF-8 by two imidazole derivatives. All isotherms exhibited a type I adsorption isotherm. The adsorption of N 2 was steeply increased at low relative pressure which suggested the existence of micropore. Table 1 shows summary data of surface area, average pore size and pore volume of the synthesized samples. In the case of exchanged ZIF-8 samples with 2-NH 2 bZIM and 2-PhIM ligands, the surface areas were enhanced when compared with the original ZIF-8. However, 2-NH 2 bZIM and 2-PhIM were bulkier than the original ligand (2-MeIM), thus the exchanging of those led to less accessible porosity due to the blocking of the exchanged linkers. In addition, the increase in surface area upon exchanging with the new ligands might be due to the removal of trapped TMPAB from the pores during the exchanged process. Therefore, to prove this assumption, the blank experiment was conducted by exchanging ZIF-8 by the same procedure without adding the imidazole derivatives. It was found that the surface area of blank sample was 1470.1 m 2 /g. This indicated that increase in surface area upon exchanging with imidazole derivatives was due to the removal of some TMPAB from the pores.  Figure 16 shows the CO2 uptake of ZIF-8 and two ZIF-8 exchanged samples obtained from the water based synthesis method, and ZIF-8 obtained from the solvothermal method. It is clearly seen that upon exchanging the 2-MeIM linker with imidazole derivatives, the CO2 uptake was enhanced. In order to explain the CO2 uptake ability of synthesized ZIF-8 by hydrogen bond acceptor assisted and linker exchanged samples, data analysis of the surface properties for those samples were performed. Figure 17 shows N2 adsorption desorption of original ZIF-8 (with the ratio of 8:2) and exchanged ZIF-8 by two imidazole derivatives. All isotherms exhibited a type I adsorption isotherm. The adsorption of N2 was steeply increased at low relative pressure which suggested the existence of micropore. Table 1 shows summary data of surface area, average pore size and pore volume of the the pores during the exchanged process. Therefore, to prove this assumption, the blank experiment was conducted by exchanging ZIF-8 by the same procedure without adding the imidazole derivatives. It was found that the surface area of blank sample was 1470.1 m 2 /g. This indicated that the increase in surface area upon exchanging with imidazole derivatives was due to the removal of some TMPAB from the pores.  Between ZIF-8 (8:2) and ZIF-8-blank samples, based on the assumption that TMPAB was completely removed from the samples, the enhancing of CO 2 uptake may be due to the increasing of internal surface area. For the 2-NH 2 bZIM exchanged sample, it exhibited a higher CO 2 uptake than that of the original ZIF-8 and ZIF-8-blank samples ( Figure 16). As describe above that higher surface area of exchanged ZIF-8 samples were due to the removal of some TMPAB from the sample during the exchanged process, and consequently the amount of CO 2 uptake of exchanged samples should be lower than ZIF-8-blank sample. Therefore, higher CO 2 uptake of exchanged samples indicated that increase in surface area was not the main factor to dominate the CO 2 uptake. This enhancing of CO 2 uptake was probably due to some additional interaction between the exchanged linker and some TMPAB with CO 2 . The CO 2 uptake of the 2-NH 2 bZIM exchanged sample was higher than that of 2-PhIM-exchanged and ZIF-8-blank samples, due to a higher amount of exchanged 2-NH 2 bZIM (around 11% from 1 HNMR analysis). This enhancing of CO 2 uptake was due to additional interaction between CO 2 and the amine group of 2-NH 2 bZIM. The increased in CO 2 uptake can be attributed to two factors. First, the basic amino groups have a high affinity toward an acidic CO 2 molecule. These results are in agreement with L. Xiang et al. [42], who reported the synthesized amino-functionalization by the in-situ substitution of 2-NH 2 bZIM into the ZIF-7. The results show that the CO 2 uptake of the amino-functionalization sample was higher than that of pristine ZIF-8. They proposed that the higher CO 2 uptake was probably due to the CO 2 affinity of the polar amino group on amino-modified ZIF-8 framework. The second factor was the quadrupole-π electron interaction between the CO 2 molecule and the imidazole and/or phenyl ring in 2-NH 2 bZIM, which was strengthened by the electron-donating group (amino functional group) [43]. Accordingly, the TMPAB in all samples may also promote the CO 2 uptake by the interaction between the delocalized π-aromatic system of the phenyl group and CO 2 . In contrast, the lower exchanged amount of 2-PhIM (around 4% from 1 H-NMR analysis) led to nearly the same amount of CO 2 uptake to ZIF-8-blank sample, which may be due to higher surface area of this exchanged sample being dominant in the CO 2 uptake behavior.

Conclusions
In this work, a zeolitic imidazolate framework-8 (ZIF-8) was synthesized by the water based synthesis method with hydrogen bond acceptor assisted and modified ZIF-8 by solvent-assisted ligand exchange method to enhance the CO 2 adsorption performance. Several synthesis conditions were investigated. The effects of mixed ordering steps of precursors showed that the mixing of 2-MeIM and quaternary ammonium salts (QAS) before adding Zn(NO 3 ) 2 ·6H 2 O solution provided a ZIF-8 with high crystallinity. The quaternary ammonium salts were varied by using three different compounds: tetraethylammonium bromide (TEAB), tetrabutylammonium bromide (TBAB) and trimethyl phenyl ammonium bromide (TMPAB). The results show that ZIF-8 synthesized by forming in TMPAB exhibited good size distribution and the smallest particle sizes compared with two other QAS. This can be explained by the higher hydrogen bonding strength of 2-MeIM and TMPAB, which led to an increase in the 2-MeIM deprotonation rate, which then sped up the coordination of linker with Zn 2+ ion. This effect can lead to a high nucleation rate and crystal growth of ZIF-8, which also led to uniformly distributed ZIF-8 particles. Moreover, the effects of various ratios between 2-MeIM and TMPAB were also studied. The ratio of 8:2 provided a good distribution of ZIF-8 particles and the smallest particle size was also obtained. Therefore, the optimal synthesis conditions were using TMPAB as a QAS with a ratio between 2-MeIM and TMPAB of 8:2 and when TMPAB was first mixed with 2-MeIM before the Zn(NO 3 ) 2 ·6H 2 O solution was added. The synthesized ZIF-8 was then modified by solvent-assisted ligand exchange methods to enhance the CO 2 uptake. The 2-MeIM linker in ZIF-8 was exchanged by two imidazole derivatives: 2-aminobenzimidazole (2-NH 2 bZIM) and 2-phenylimidazole (2-PhIM). The CO 2 uptake was carried out by thermogravimetric analysis (TGA) at atmospheric pressure. The results show that exchanged ZIF-8 samples with 2-NH 2 bZIM exhibited higher CO 2 uptake than that of pristine ZIF-8 due to two factors; i.e., (i) an additional interaction between CO 2 and exchanged imidazole derivatives and (ii) increasing surface area, pore size and pore volume upon exchanging with the imidazole derivative.
Author Contributions: The present research project has been initiated by K.K. and K.C.C.; sample preparation and characterization by K.K.; writing-original draft preparation by K.C.C.; 1 H-NMR interpretation by S.M. and P.J. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.