Design and synthesis of metal organic frameworks doctor of philosophy major chemical engineering

Therefore, The development of a CO-selective adsorbent with large CO adsorption capacity, high CO/CO2 selectivity, and good stability is still a huge challenge.. C7H8 is relatively diver

Dissertation for the Degree of Doctor of Philosophy

Design and Synthesis of Metal-Organic Frameworks

for CO, CO2, and C7H8 Adsorption

Le Van Nhieu

Department of Chemical Engineering

Graduate School Kyung Hee University Seoul, Korea

June, 2021

Design and Synthesis of Metal-Organic Frameworks

for CO, CO2, and C7H8 Adsorption

Le Van Nhieu

Department of Chemical Engineering

Graduate School Kyung Hee University Seoul, Korea

June, 2021

Design and Synthesis of Metal-Organic Frameworks

for CO, CO2, and C7H8 Adsorption

by

Le Van Nhieu

Advised by

Prof Jinsoo Kim

Submitted to the Department of Chemical Engineering and the Faculty of the Gradual School of

Kyung Hee University in partial fulfillment

of the requirement for degree of Doctor of Philosophy

Dissertation Committee Chairman Prof Eun Yeol Lee

Prof Bum Jun Park Prof Chang Kyoo Yoo Prof Kye Sang Yoo Prof Jinsoo Kim

The separation of CO out of gas mixture, especially containing CO2 is an important mission in the industrial production sector but encounter huge challenges due to the higher polarizability of CO2 than that of CO Most of the investigation showed that after introducing Cu(I) into pore system of MOF-support, the resulting materials exhibited a higher adsorption capacity of CO than CO2 whereas a contrary result was observed on the original MOFs This

is due to -complexation formed between Cu(I) and CO species Among the reported MOFs, MIL-100(Fe) possesses high BET surface area, thermal stability ( ̴ 320 oC), and tunability

of the oxidation state of iron ions (Fe(II) and Fe(III)) under high temperature (150 ̴ 250 oC)

So, a simple route is employed to introduce Cu(I) on MIL-100(Fe), in which Cu(II) is directly transferred to Cu(I) thanks to Fe(II), but no requirement support from reducing agents However, MIL-100(Fe) is typically synthesized in closed batch systems, which is

not favorable for large-scale production Herein, we report a scalable MOF synthesis route based on a continuous flow tubular reactor equipped with microwave volumetric heating The system enabled continuous crystallization of MIL-100(Fe) with a high space-time yield

of ~771.6 kg m-3 day-1 under relatively mild conditions in a range of temperature (100 ̴ 110

oC) and resident time of 50 min The product quality is evaluated via porous property and crystallinity in comparison to the traditional method Ultimately, the MIL-100(Fe) was used

as a support to prepare Cu(I)-modified π complexation adsorbents The adsorbents exhibited preferred CO adsorption over CO2, and the adsorption performance was confronted to, or even higher than most of the Cu(I)-modified π complexation adsorbents in previous reports Until now, the CO-selective adsorbents are kept developing towards the improvement

of CO uptake capacity and CO/CO2 selectivity, but Cu(I)-incorporated MOFs are instability

in the air This is the main reason for reducing CO separation performance in the real gas environment being usually available a certain amount of oxygen and moisture Recently, some reports have revealed a strategy to improve the stability of Cu(I)-incorporated MOFs, however, their CO adsorption capacity is modest Therefore, The development of a CO-selective adsorbent with large CO adsorption capacity, high CO/CO2 selectivity, and good stability is still a huge challenge In this dissertation, a novel Cu(I)-incorporated MIL-100(Fe) adsorbent for CO/CO2 separation is prepared using a host–guest redox strategy by combining the co-addition of Zn(II) and Cu(II) inside the MIL-100(Fe)’s pore system The addition of Zn(II) resulted in a higher Cu(I) yield of the adsorbent due to the facilitated regeneration of Fe(II), which was utilized for the reduction of Cu(II) Therefore, both CO uptake amount and achieved CO/CO2 selectivity on Cu(I)Zn@MIL-100(Fe) with only 10 wt% of Zn loading were considerably higher than that of the benchmark Cu(I)-incorporated adsorbents In addition, the presence of the Zn(II) in Cu(I)Zn@MIL-100(Fe)-10 improved the oxygen resistance This study opens a new perspective for developing efficient CO-selective π-complexation adsorbents with high CO/CO2 selectivity and superior oxygen resistance

Unlike CO adsorption, the MOF-adsorbent for capturing the target gas like CO or

C7H8 is relatively diverse, in which the adsorbent perhaps possesses positive factors for gas adsorption like a superior surface area, a suitable pore structure, and a large amount of adsorption sites having an affinity toward adsorbates Zirconium-based MOFs (UiO-66, UiO-67) are potential adsorbents for gas adsorption due to a quite large surface area, easily tunable pore structure as well as chemical surface, high chemical/thermal stability, and facilely large scale production thanks to using a microwave-assisted continuous tubular reactor However, they exhibited a modest uptake capacity for both CO2 and C7H8 in comparison with the others So, gas adsorption capacity should be improved

For CO2 adsorption, an amino-defective UiO-66 was prepared by a one-step synthesis method using the mixed linkers of terephthalic acid and a cheap defect linker as 4-aminobenzoic acid The presence of the 4-aminobenzoic acid in the reaction system, induced enhanced porosity owing to the missing-linker defects, simultaneously, created the amino (-

NH2) groups in the framework Both two factors contribute to the improvement of the CO2

capture capacity on modified UiO-66 as a result of the synergy effect Only with 10% in mole of used 4-amino benzoic acid in the mixed ligand, at 25 oC and 1 bar, the obtained CO2

uptake amount and ideal adsorbed solution theory-based CO2/N2 selectivity on the resulting material increased 47.8% and afforded 2.8 higher times in comparison with the original UiO-

66 sample, respectively These results exhibited an efficient approach (a cheap linker as aminobenzoic acid and one-step synthesis) to prepare a defective UiO-66 adsorbent with amine functional groups, which not only improve CO2 separation performance but also reduce production cost

4-For C7H8 adsorption, some defective Zr-based biphenyl dicarboxylate (UiO-67) MOFs were prepared via fast modulated synthesis under microwave-assisted continuous tubular reactor by using formic, acetic, propionic, and benzoic acid as modulators A surface-modified UiO-67(Zr) framework with high porosity and crystallinity could be rapidly produced in a few minutes due to the incomplete exchange between the bridging ligand and the modulator The defect concentration in the products was tuned by controlling both the modulator species and their concentrations The adsorption ability toward toluene of the

prepared UiO-67(Zr) MOFs was found to be related to their structural defects The defective UiO-67(Zr) MOF synthesized with HCOOH as the modulator exhibited the highest toluene adsorption capacity (467 mg g –1), surpassing also most of the previously reported adsorbent materials, such as zeolites, activated carbon, UiO-66(Zr), H2N-UiO-66(Zr), ZIF-67, and CuBTC Moreover, the experimental dynamic adsorption data were mathematically modeled

to predict the adsorption behaviors of defective UiO-67(Zr) MOFs

Additionally, zirconium-based MOFs has still had limitations in fixed-bed adsorption system owing to its tiny sub-micron crystallite size leading to inconvenience in transportation, difficult recovery and serious pressure loss in fixed-bed adsorption system Thus, an approach to construct mm-scale granules using UiO-66(Zr) powder is required In this work, UiO-66(Zr) particles were prepared by the solvothermal method under microwave irradiation for only 20 min, then fabricated into spherical granules of UiO-66/PVA by freeze granulation technique PVA was added as a binder to connect UiO-66 particles together to spherical beads with high mechanical strength, not affecting the crystalline and micropore structures of UiO-66 The regular octahedron of the UiO-66 individual particles remained intact and the pore size did not change with increasing PVA concentration However, PVA bound the particles together to form compact and cohesive network clusters that reduced the BET surface area and total pore volume This consequently lowered the toluene adsorption efficacy slightly due to the premature breakthrough that limited the toluene molecules exposure into the micropores of the individual UiO-66 particles

Table of Contents

ABSTRACT i

List of Tables ix

List of Figures xi

CHAPTER 1 - Introduction 1

1.1 Background 1

1.2 Motivation 8

1.3 Research objectives 10

1.4 Dissertation overview 12

CHAPTER 2 – Literature Review 15

2.1 Adsorption technique 15

2.1.1 Adsorption definition 15

2.1.2 Adsorption models 18

2.1.3 Isosteric heat of adsorption 19

2.1.4 Ideal adsorption solution theory (IAST) selectivity 20

2.2 Metal organic frameworks 22

2.2.1 Iron-based metal organic framework (MIL-100Fe) 26

2.2.2 Zirconium-based metal organic framework (UiO-66 & UiO-67) 28

2.3 Application of MOFs for gas adsorption 30

2.3.1 Mechanism of gas adsorption 30

2.3.2 Metal organic frameworks for CO adsorption 33

2.3.3 Metal organic framework for CO2 adsorption 36

2.3.4 Metal organic framework for toluene adsorption 37

CHAPTER 3 -Microwave-assisted continuous flow synthesis of MIL-100 (Fe) and its application to Cu(I)-loaded adsorbent for CO/CO 2 separation 40

3.1 Introduction 40

3.2 Materials and methods 41

3.2.1 Chemicals 41

3.2.2 Synthetic procedures 41

3.2.3 Characterization 42

3.3 Results and discussion 43

3.3.1 Synthesis of MIL-100(Fe) in a microwave-assisted flow reactor 43

3.4 Conclusions 61

CHAPTER 4 -A novel approach to prepare Cu(I)Zn@MIL-100(Fe) adsorbent with high CO adsorption capacity, CO/CO 2 selectivity and stability 62

4.1 Introduction 62

4.2 Experimental 64

4.2.1 Materials 64

4.2.2 Synthesis of MIL-100(Fe) 64

4.2.3 Synthesis of Cu(I)@MIL-100(Fe) and Cu(I)Zn@MIL-100(Fe) adsorbents 64

4.2.4 Characterizations 65

4.2.5 CO and CO2 adsorption test 65

4.3 Results and discussion 68

4.3.1 Characterizations of MIL-100(Fe) and Cu(I)Zn@MIL-100(Fe) 68

4.3.2 CO and CO2 adsorption on Cu(I)Zn@MIL-100(Fe) adsorbents 75

4.3.3 Regeneration and stability test 84

4.3.4 Reduction mechanism 86

4.4 Conclusions 88

CHAPTER 5 - Facile one-step synthesis of amino-defective UiO-66 using 4-amino benzoic acid for enhanced CO 2 adsorption performance 90

5.1 Introduction 90

5.2 Experimental 93

5.2.1 Materials 93

5.2.2 Synthesis of MOF materials 93

5.2.3 Characterizations 93

5.2.4 CO2 and N2 adsorption 94

5.3 Results and Discussions 94

5.3.1 Material characterizations 94

5.3.2 CO2 and N2 adsorption 110

5.3.3 Isosteric heat of CO2 adsorption and the regeneration of the adsorbent 114

5.3.4 Adsorption selectivity of CO2/N2 on UiO-66 and UiO-66#10-NH2 118

5.4 Conclusions 119

CHAPTER 6 - Defect engineering of UiO-67(Zr) under continuous-flow microwave synthesis condition and application for toluene adsorption 120

6.1 Introduction 120

6.2 Experimental 122

6.2.1 Synthesis procedure 122

6.2.2 Characterization 123

6.2.3 Toluene recovery test 125

6.3 Results and discussion 125

6.3.1 Microwave-assisted continuous-flow synthesis of UiO-67(Zr) 125

6.3.2 Adsorption and desorption of toluene 142

6.3.3 Dynamic adsorption of toluene 146

6.4 Conclusions 148

CHAPTER 7 - Facile synthesis of UiO-66/PVA spherical granules and their application for toluene adsorption 149

7.1 Introduction 149

7.2 Experimental 152

7.2.1 Materials 152

7.2.2 Synthesis of UiO-66(Zr) assisted by microwave and UiO-66/PVA beads 152

7.2.3 Characterizations 155

7.2.4 Toluene adsorption/desorption 155

7.3 Results and discussion 156

7.3.1 Characterizations of UiO-66 and UiO-66/PVA granules 156

7.3.2 Toluene adsorption/desorption test 165

7.4 Conclusions 174

CHAPTER 8 - Summary and further works 175

8.1 Summary 175

8.2 Further works 178

Acknowledgements 180

References 181

List of Tables Table 3.1 Texture properties of MIL-100(Fe) synthesized in a microwave assisted

continuous flow tubular reactor (MW-MIL-100(Fe)), synthesized in a conventional batch reactor (CB-MIL-100(Fe)), and the corresponding data previously reported in the literature

*: Brunauer–Emmett–Teller (BET) surface area, #: specific pore volume, $: assisted flow tubular reactor, and $$: conventional batch reactor 46

microwave-Table 3.2 Comparison of MIL-100(Fe) yields synthesized in a microwave-assisted

continuous flow tubular reactor (MW-MIL-100(Fe)), synthesized in a conventional batch reactor (CB-MIL-100(Fe)), and the corresponding data previously reported in the literature ($: microwave assisted flow tubular reactor, $$: conventional batch reactor, and : The obtained yield of the product with no purification) 50

Table 3.3 Texture properties of MIL-100(Fe) synthesized in a microwave assisted

continuous flow tubular reactor (MW-MIL-100(Fe)) as a function of CuCl loading (*: BET surface area and #: specific pore volume) 51

Table 3.4 Fitted dual-site Langmuir–Freundlich (DSLF) parameters for CO and CO2

isotherms experimentally obtained for MW-MIL-100(Fe) and xCuCl@MIL-100(Fe) at 298

K 57

Table 3.5 Comparison of equilibrium CO adsorption performances of xCuCl@MIL-100(Fe)

with reported Cu(I)-modified π complexation adsorbents Equilibrium selectivity was calculated by taking ratio between CO and CO2 adsorption capacity at 100 kPa and 298 K 59

Table 4.1 Textural properties of MIL-100(Fe), Cu(I)@MIL-100(Fe), and

PABA concentration 105

Table 5.2 Components of oxygen calculated from O 1s spectra 105

Table 5.3 Comparison of CO2 uptake capacity on different adsorbents at 1bar 112

Table 5.4 The obtained parameters and correlation coefficients from fitting the Langmuir– Freundlich model 117

Table 6.1 Microwave –assisted continuous flow synthesis of UiO-67(Zr) 127

Table 6.2 Number of linkers per Zr6 formular unit and linker deficiency derived from TGA analyses 139

Table 6.3 Equilibrium toluene adsorption capacity of various adsorbent materials 144

Table 6.4 Adsorption parameters of the Yan and Thomas models 147

Table 7.1 Effect of PVA content on compressive strength 157

Table 7.2 BET surface area and pore volume of UiO-66 and UiO-66/PVA samples 162

Table 7.3 Toluene uptake capacity of various MOF-adsorbents 168

Table 7.4 The model fitting parameters 170

List of Figures Figure 2.1 Three-step gas adsorption Reproduced with ratification from ref [90] Copyright

2020 Elsevier 17

Figure 2.2 Several representative MOFs Reproduced with ratification from ref [99]

Copyright 2013 Springer Nature 25

Figure 2.3 Components of flow chemistry Reprinted with ratification from ref [65]

Copyright 2017 Elsevier 25

Figure 2.4 Formation of MIL-100(Fe) structure Reprinted with ratification from ref [115]

Copyright 2012 John Wiley and Sons 27

Figure 2.5 Effect of activation temperature on the oxidation state of iron (open metal sites)

in MIL-100(Fe) structure Reprinted with ratification from ref [113] Copyright 2019 Royal Society of Chemistry (Great Britain) 27

Figure 2.6 Zirconium-based UiO metal organic framework: (a) UiO-66 with benzen

dicarboxylate, (b) UiO-67 with biphenyl dicarboxylate, (c) UiO-68 with terphenyl dicarboxylate Reproduced with ratification from ref [119] Copyright 2008 American Chemical Society 29

Figure 2.7 Nucleus structures of UiO’s cluster Zirconium – red, oxygen – blue, carbon –

grey and hydrogen – white Reprinted with ratification from ref [119] Copyright 2008 American Chemical Society 29

Figure 2.8 The diffusion models and effect of pore size on diffusion coefficient Reprinted

with ratification from ref [124] Copyright 2020 Elsevier 31

Figure 2.9 Formation of π-complexation between CO and metal atom Reproduced with

ratification from ref [128] Copyright 2010-2021 Atlanta Publishing House LLC 35

Figure 3.1 Schematic illustration of a continuous flow reactor combined with a microwave

oven 45

Figure 3.2 Electron micrographs of MW-MIL-100(Fe) at varying temperature of (a) 353 K,

(b) 363 K, (c) 373 K, and (d) 383 K during fixed residence time (50 min) and during varying residence time of (g) 20 min, (h) 30 min, (i) 40 min, and (j) 50 min at fixed temperature (373 K); corresponding XRD patterns of MW-MIL-100(Fe) synthesized (e) at varying temperature and (k) during varying residence time; corresponding N2 isotherms at 77 K of MW-MIL-100 (Fe) synthesized (f) at varying temperature and (l) during varying residence time MW and CB in the XRD and N2 isotherm figures indicates microwave-assisted flow tubular reactor and conventional batch reactor, respectively 47

Figure 3.3 FE-SEM images of MIL-100(Fe) synthesized in a conventional batch reactor

(CB-MIL-100(Fe)) at (a) 383 K for 50 min, (b) 383 K for 24 h, (c) 373 K for 50 min, and (d) 373 K for 24 h 48

Figure 3.4 (a) electron micrograph, and (b) Fe and (c) Cu elemental maps of

45CuCl@MIL-100(Fe); (d) XPS spectra of 45CuCl@MIL-100(Fe), 45Cu(II)@MIL-100(Fe), CuCl, and CuCl2; (e) XRD patterns of MW-MIL-100(Fe), 45CuCl@MIL-100(Fe), CuCl, and CuCl2 53

Figure 3.5 (a) CO and (b) CO2 isotherms at 298 K, and (c) corresponding ideal adsorbed solution theory (IAST)-predicted CO/CO2 binary adsorption selectivities of xCuCl@MIL-100(Fe) for an equimolar CO/CO2 mixture 54

Figure 3.6 IAST predicted CO/CO2 binary adsorption isotherms at 298 K for an equimolar CO/CO2 mixture of (a) MW-MIL-100(Fe), (b) 10CuCl@MIL-100(Fe), (c) 20CuCl@MIL-100(Fe), (d) 30CuCl@MIL-100(Fe), (e) 40CuCl@MIL-100, and (f) 45CuCl@MIL-100(Fe) 55

Figure 4.1 PXRD patterns of MIL-100(Fe), Cu(I)@MIL-100(Fe), and

Cu(I)Zn@MIL-100(Fe)-12 with referenced peaks of ZnCl2, CuCl2, and CuCl 67

Figure 4.2 (a) SEM image of Cu(I)Zn@MIL-100(Fe)-10 and EDX mapping of (b) Fe, (c)

Cu, and (d) Zn 67

Figure 4.3 (a) N2 adsorption and desorption isotherms at 77 K and (b) pore size distributions (Horvath-Kawazoe model) for MIL-100(Fe), Cu(I)@MIL-100(Fe), and Cu(I)Zn@MIL-

100(Fe)-x samples 69

Figure 4.4 (a) Thermogravimetric curves for MIL-100(Fe), Cu(I

100(Fe)-10 samples (b) FT-IR spectra of MIL-100(Fe)-100(Fe), Cu(I)@MIL-100(Fe)-100(Fe) and

Cu(I)Zn@MIL-100(Fe)-x samples 73

Figure 4.5 Cu 2p core-level XPS spectra of (a) Cu(I)@MIL-100(Fe), (b)

Cu(I)Zn@MIL-100(Fe)-5, (c) Cu(I)Zn@MIL-100(Fe)-10, (d) Cu(I)Zn@MIL-100(Fe)-12 and (e) CuCl2, respectively 74

Figure 4.6 Gas adsorption isotherms on the adsorbents at 298 K: (a) CO, (b) CO2 77

Figure 4.7 Comparison of CO/CO2 separation factor as a function of CO working capacity for Cu(I)Zn@MIL-100(Fe)-10 and the benchmark Cu(I)-containing adsorbents 78

Figure 4.8 (a) IAST-predicted CO/CO2 selectivity for equimolar CO/CO2 mixture (50:50)

at 25 oC on various Cu(I)Zn@100 adsorbents (b) Enthalpy of CO adsorption on

MIL-100(Fe), Cu(I)@MIL-MIL-100(Fe), and Cu(I)Zn@MIL-100(Fe)-10 adsorbents 82

Figure 4.9 CO adsorption/desorption isotherms on Cu(I)Zn@MIL-100(Fe)-10 at 298 K for

5 cycles 83

Figure 4.10 Evolution of CO adsorption capacities of Cu(I)@MIL-100(Fe) and

Cu(I)Zn@MIL-100(Fe)-10 with the time exposed to the air CO adsorption capacity was collected at 298 K and 100 kPa 85

Figure 4.11 Fe 2p core–level XPS spectra on (a) as-prepared MIL-100(Fe), (b) activated

MIL-100(Fe), (c) Cu(I)@MIL-100(Fe), (d) 100(Fe)-5, (e) 100(Fe)-10, and (f) Cu(I)Zn@MIL-100(Fe)-12 86

Cu(I)Zn@MIL-Figure 4.12 Zn 2p core–level spectra of Cu(I)Zn@MIL-100(Fe)-x samples with referenced

ZnCl2 86

Figure 5.1 SEM images of of the UiO-66 and modified UiO-66 samples 96

Figure 5.2 XRD patterns of the UiO-66 and modified UiO-66 samples 97

Figure 5.3 1H-NMR spectra of the UiO-66 and modified UiO-66 samples 97

Figure 5.4 FT-IR profiles of the as-prepared materials 100

Figure 5.5 TGA curves of the as-prepared materials 102

Figure 5.6 (a) N2 adsorption/desorption isotherms at 77 K, and (b) pore size distribution of the as-prepared materials, as calculated from the HK model 103

Figure 5.7 XPS spectra of the as-prepared materials: (a) elemental surveys, (b) C 1s region, (c) Zr 3d region, (d) O 1s region, and (e) N 1s region 108

Figure 5.8 The deconvoluted O 1s spectra of the as-prepared materials 109

Figure 5.9 CO2 and N2 adsorption properties of the UiO-66 and modified UiO-66 samples: (a) CO2 adsorption isotherms, (b) N2 adsorption isotherms, and (c) CO2 adsorption capacities based on BET surface area 113

Figure 5.10 CO2 isotherms at several different temperature for (a) 66 and (b) UiO-66#10-NH2 115

Figure 5.11 Isosteric heat of CO2 adsorption on UiO-66 and UiO-66#10-NH2 115

Figure 5.12 Isotherms for the CO2 adsorption/desorption on UiO-66#10-NH2 at 298 K for five cycles 116

Figure 5.13 IAST-predicted selectivities using CO2/N2 (0.15/0.85) mixture on UiO-66 and UiO-66#10-NH2 at 298 K 116

Figure 6.1 Synthetic procedure (DMF: N,N’-dimethylformamide; BPDC: 4,4’-biphenyl dicarboxylic acid, UiO-67(Zr): Zr-based biphenyl dicarboxylate) 124

Figure 6.2 XRD patterns of the prepared UiO-67(Zr) samples with (a) different modulators

and (b) different modulator concentrations 128

Figure 6.3 FT –IR spectra of defecive UiO-67(Zr) prepared with diffent (A) Modulators: (a)

UiO-67(Zr)-HFo, (b) UiO-67(Zr)-HAc, (c) UiO-67(Zr)-HFo; (B) Modulator concentrations:

(a) UiO-67(Zr)-HFo-40eq., (b) UiO-67(Zr)-HFo-50eq., (c) UiO-67(Zr)-HFo-60eq., and (d) UiO-67(Zr)-HFo-80 eq 131

Figure 6.4 SEM images of (a) UiO-67(Zr)–HFo, (b) UiO-67(Zr)–HAc, (c) UiO-67(Zr)–HPr.

Figure 6.7 XPS analyses of UiO-67(Zr) with different modulators ((a), (b), (c)), and

HCOOH concentrations ((d), (e), (f)) 137

Figure 6.8 TGA curves of the defective UiO-67(Zr) samples with different HCOOH

amounts: (a) UiO-67(Zr) -40 eq., (b) UiO-67(Zr) -50 eq., UiO-67(Zr) -60 eq., and 67(Zr) -80 eq 140

(d)UiO-Figure 6.9 Breakthrough curves and toluene adsorption capacity on defective UiO-67(Zr)

samples prepared with different (a, b) Modulators and (c, d) Modulator concentrations 141

Figure 6.10 Breakthrough curves of toluene on defective UiO-67(Zr) –HFo adsorbent at

different temperatures 145

Figure 6.11 Consecutive adsorption–desorption cycles over generated defective

UiO-67(Zr)-HFo sample at 25 oC and atmospheric pressure 145

Figure 7.1 Preparation of UiO-66/PVA spherical granules from UiO-66(Zr) powder 154

Figure 7.2 Spherical granules of UiO-66/PVA-90 in different sizes based on the needle’s

type 157

Figure 7.3 FE-SEM images of UiO-66 (Zr) and UiO-66/PVA samples 158 Figure 7.4 XRD patterns of UiO-66(Zr) and UiO-66/PVA samples 158 Figure 7.5 Pore analyses of UiO-66 and UiO-66/PVA samples (a) N2 adsorption/desorption isotherms, (b) pore size distributions calculated by H-K method 161

Figure 7.6 FTIR spectra of UiO-66 and UiO-66/PVA with different PVA contents 164 Figure 7.7 TGA curves of UiO-66 particles and UiO-66/PVA granules with different PVA

contents 164

Figure 7.8 Experimental toluene breakthrough and the fitted Yoon-Nelson model 167 Figure 7.9 Toluene adsorption-desorption cycles at 295 K: (a) UiO-66, (b) UiO-66/PVA-95,

and (c) UiO-66/PVA-90 and (d) UiO-66/PVA-80 172

Figure 7.10 Toluene adsorption amounts on UiO-66 particles and UiO-66/PVA granules at

295 K for three cycles For second and third cycle after desorption at 423 K under vacuum during 2 h Dark yellow – first cycle; blue – second cycle; olive – third cycle 173

to climate change and negatively affect on the ecosystems, consequently threatened the health and life of human [1-4] Evidence for this hazard, the World Health Organization (WHO) released that there were around 4.2 million deaths of young people in 2016, primarily in developing countries, originated from poor quality of the air [5] In addition, some pollutants are feedstock in chemical production processes For instance, carbon monoxide (CO) is widely used as a valuable raw material for producing methanol, ethanol, acetic acid, plastics, and fibers [6-8] Carbon dioxide (CO2) is a raw material for the synthesis

of urea, polycarbonate, methanol, ethanol, fertilizer… [9, 10] So, in the 21st century, humanity must face an extremely huge challenge relating to habitat and search for suitable solutions to mitigate air pollution and utilize available raw material sources in the air

In order to obtain this purpose, some efforts in using renewable energy to replace traditional ones, consuming efficiently energy by utilizing waste energy sources and improving fuel efficiency, using recycling materials, etc are practical solutions contributing

to the declining volume of the air pollutants emitted into the atmosphere from the industrial sector as well as anthropogenic activities [5, 11] Nevertheless, this approach is not feasible, especially for developing countries because of incompatible infrastructure, leading to a significant increment of overall production cost [10, 12] Therefore, technological resolutions having the ability to control the concentration of the gas pollutants in the exhaust gas should be orientated to application because they are easy to integrate into the existed

systems [10, 12] Thiscould be implemented thanks to gas separation technologies such as cryogenic condensation/distillation, absorption, adsorption, and membrane separation [10, 13-19] By the way, the captured gas pollutants are easily managed and then can be utilized

as a feedstock in various processes for the next steps

In gas separation technique, cryogenics is known as a primitive pathway used to capture gases via liquefaction of one or several components at low temperature based on their dew point temperature In the case of recovered species into liquid condensed multi-components simultaneously, a distillation technique is followed to separate species into single components There are reports applying this method for separating CO and H2 out of syngas [18], CO from natural gas [9], CO2 out of mixture of post-combustion flue gas [20, 21], toluene (C7H8) in presence of N2 [22], hydrocarbons from natural gas [23] This method has been received consideration for capturing gas, especially for VOCs because of the simple process with the appearance of physical conversion presence and no secondary pollutants generated during the process Moreover, recovered gas is in the liquid phase for advantage transportation [9, 22, 24, 25] Nevertheless, there are drawbacks of operating costs such as refrigeration, released water in feed flow due to obstruction of ice in the equipment system [19, 24] In the case of cryogenic distillation, it is unfeasible for a mixture containing species with similar boiling temperatures like CO-N2 [9, 26], leading to an increment in energy cost

In addition, it is effectively applied for inlet streams with medium or high concentration [24, 27] In another scenario, absorption is also one of the methods employed to recuperate objective gases out of exhaust gas by contacting between the gas phase and liquid phase (solvent) Depending on intrinsic solvent, the absorption process is categorized into chemical absorption occurring a reversible reaction between absorbates and solvent, and physical absorption which only has dissolve of target gases into the solvent, based on Henry’s law [9, 28] Up to now, this method has been popularly used to capture most of the gases like CO,

CO2, VOCs excepting H2 [9, 10, 18, 24, 26, 28-30], however, there are several significant limitations: (i) high cost for both material and energy consumption related to solvent refreshment, (ii) generated secondary pollutants due to evaporated and degraded solvents

during regeneration at high temperature, (iii) large size of the equipment and high corrosion rate, leading to high overall cost [9, 25, 31] Difference from absorption, the interaction between contaminated gas and porous solids (absorbents), as a result, the impurity component in the gas mixture is held within the porous structure of solids Depending on the chemical surface of adsorbents, adsorption interaction is established under chemical or physical bonding [9, 24] So far, adsorption is a familiar solution applied for gas separation for several reasons such as the working condition in a wide scope of temperature and pressure, effectiveness with a low concentration of absorbate in the gas mixture, reduction

of secondary pollution in comparison to absorption, high regenerability, easy operation Besides, there are some disadvantages such as high capital, low selectivity, energy consumption to refresh the adsorbent leading to increasing operating cost [22, 25, 27] Unlike the aforementioned methods, membrane gas separation is only one species in the inlet stream diffused through the membrane thanks to driving force being different pressure between the inlet and outlet flow So, membrane technique has emerged as a competitive solution with conventional methods (cryogenics, absorption, adsorption) owing to no requirement of phase change, absent chemical system, easy operation However, it is difficult to achieve high selectivity and permeance in a one-stage process at the same time

To overcome the drawback, a multi-step membrane process is implemented, which is one of the reasons for increasing capital and operating cost [9, 10, 24, 30, 32] Therefore, the development of membrane materials is neglected over the years [9, 33] It can be seen that each method has specific advantages and disadvantages The selected technique depends on the physical and chemical properties of the pollutant as well as its concentration, investment, and operation cost, alongside with required separation performance In general, adsorption

is considered as an efficient method with simple operation, saving cost, and high efficiency, especially for gas mixture with low concentration [8, 15, 27, 34-36]

It is illustrious that the absorbent is a key factor to decide gas adsorption performance via evaluating adsorption capacity, working capacity, selectivity, recyclability, and stability depending on its inherent properties such as channel system in material, porous structure,

chemical surface, so on [6, 37-39] Until now, the conventional porous materials like metal oxides, zeolites, activated carbons, mesoporous silica, and their functionalized-materials or modified-materials have been kept investigating to find promising adsorbents for CO, CO2, VOCs, H2, … separation out of the gas mixture For CO adsorption, the adsorbents have been fabricated by incorporating transition metal salts, mostly Cu(I) Ag(I) or Pd(II), which are active metal- binding sites that can capture CO molecules through the formation of π-complexation Among various salts, Cu(I) species have been mostly used as they are cheap and readily available [36, 38] Until now, many Cu(I)-incorporated adsorbents have been investigated to find promising adsorbents for CO/CO2 separation Nevertheless, these adsorbent materials have shown low CO adsorption capacity such as CuCl/bayerite (2.16 mmol g-1) [6], CuCl/boehmite (1.56 mmol g-1) [8], CuCl/MCM-41 (0.57 mmol g-1) [40], CuCl/γ-Al2O3 (1.0 mmol g-1) [41], and CuZSM-5 (0.11 mmol g-1) [42] and/or low CO/CO2

or CO/N2 selectivity like CuCl/AC [43, 44], CuCl/Y [45] Difference from CO-selective adsorbents, adsorption interaction between CO2 or toluene (C7H8) as a compound represented for VOC depends on the surface area as well as a reasonable pore structure contributing to the improvement of van der Waals force Moreover, surface modification on the structure of adsorbents with the presence of functional groups like amino (-NH2), hydroxy (OH), carboxylic (COOH), sulfonate (HSO3) or metal oxide, unsaturated metal sites contributes to the improvement of uptake amount and selectivity due to their strong affinity to adsorbates (CO2, C7H8) [46-49] However, the similar results to CO-selective adsorbents have been also observed on these adsorbents with low adsorption capacity for

CO2 (MCM-41 (0.79 mmol g-1) and Fe/MCM-41 (0.87 mmol g-1) [46], ZSM-5 (1.5 mmol g1

-), Y (1.25 mmol g-1) [50], activated carbon (1.8 mmol g-1) [51], ACs (1.84 mmol g-1), ACs (2,2 mmol g-1) [52], Si-MCM-41 (0.62 mmol g-1), Si-MCM-41-PEI-50 (0.75 mmol g-1) [53]) and toluene (Zeolite (30.7 mg g-1) [54], lignin-based activated carbon (169.4 mg g-1) [55], AC (41 mg g-1), ZnO@AC (68 mg g-1) [49], SiO2 (5.2 mg g-1), PDVB/R-SiO2-0.5 (61

NiO-mg g-1)) [56] This is due to the low surface area of conventional absorbents In addition, it

is difficult to tune the pore structure and modify the chemical surface of the traditional absorbents [39, 57, 58] Therefore, it is needful to develop novel materials having high

adsorption capacity, selectivity, and regenerability

Luckily, Metal organic frameworks (MOFs) are inorganicorganic hybrid materials constructed by combining metal ions or clusters and organic linkers Unlike conventional adsorbents, MOFs hold outstanding properties such as large total pore volume, high specific surface area, thermal stability and can easily adapt chemical surface and porosity [15, 35,

59, 60] Moreover, up to now, thousands of MOFs have been fabricated and developed by many means and different approaches Traditionally, most of the MOFs are prepared via solvothermal batch reaction in Teflon-lined stainless steel bomb reactors, which requires very long reaction times (from 12 h to a few days), with high energy consumption, and produces only a small MOF amount Microwave–assisted synthesis has recently emerged as

an energy–effective MOF preparation approach due to its highly efficient heat transfer and short reaction time compared with conventional heating methods [61-63]

On the other hand, alongside with batch reaction in Teflon-lined stainless steel bomb reactors, some different methods such as mechanochemistry, electrochemistry, and flow chemistry are employed to synthesize MOFs [64, 65] Among them, the flow chemistry has gradually become a nucleus method possessing tremendous strength which can be mentioned such as improvement of efficient heat and mass transfer, leading to shortening reaction time, and consequently to reduce production cost [65] More interestingly, this method is easy to large scale, leading to increasing productivity [64, 66] Kim et al [67] fabricated Cu-BTC by employing a continuous tubular reactor made of stainless steel with

30 cm of length and 1.59 mm of inside diameter and using an electric heating source The reaction was operated at a temperature of 120 oC and a reaction time of 5 min Interestingly, the obtained quality product (crystallinity, BET surface area, and morphology) was similar

to that prepared by the conventional method with the same recipe and reaction temperature during 18h In another exhibition, Colin McKinstry and colleagues [68] successfully designed a system of continuously stirred tank reactors to synthesize MOF-5 The reaction was carried out at a temperature of 140 oC, resident time of 4 h under atmospheric pressure with a high space-time yield of 1000 kg m-3 day-1 In order to increase the throughput of

production, the Al-Fumarate MOF was successfully scaled up using continuous tubular reactors made of stainless steel with different volume (10 mL; 107 mL; 374 mL; and 1394 mL) The reaction was run at 65 oC with the resident time of 1 min and pressure of 7 bar The highest production rate and space-time yield were reached 5.6 kg h-1 and 97,159 kg m-3day-1 for obtained Al-Fumarate from the reactor of 1394 mL [66] Therefore, Using the microwave-assisted flow chemistry method for preparing MOFs becomes extremely efficient because of the integrated advantages of both flow chemistry and microwave heating The UiO-66, MIL-53(Al), and HKUST-1 were produced by using a microwave-assisted continuous tubular reactor in a short time of 7 min, 4 min, and 1 min, respectively [69] The obtained product was in large amount with high yield and space-time yield without affecting

on characteristic materials These demonstrated that increasing productivity of MOFs can be implemented easily, which meets one of the prerequisites to apply MOFs on an industrial scale

Recently, Cu(I)-incorporated MOF adsorbents, such as Cu(I)@MIL-100(Fe) [36, 39, 70] and Cu(I)@MIL -101(Cr) [71] have been reported to exhibit high CO adsorption capacity as well as high CO/CO2, CO/N2, and CO/CH4 selectivities In general, these approaches have attempted to construct the Cu(I) sites inside the pores of MOFs with high reduction yield and high dispersion of Cu(I), which remarkably enhance CO adsorption in comparison with traditional absorbents However, for practical applications, Cu(I)-based π-complexation adsorbent requires not only high CO adsorption performance but also good air stability due to the possibility of Cu(I) oxidation Very recently, Yin et al [71] synthesized CuV-incorporated MIL-101(Cr), which showed good CO selectivity and remarkable stability under exposure to atmospheric air due to the assistance of vanadium species Nevertheless, the prepared CuV@MIL-101(Cr) showed modest CO adsorption capacity (1.3 mmol g-1), which was much lower than those of other Cu(I)-incorporated MOF adsorbents Therefore, there is still a large field to develop a CO-selective MOF adsorbent possessing a large CO uptake capacity, high CO selectivity, and well-air stability

Among the thousands of MOFs, the UiO family (UiO = University of Olso has

emerged as a potential candidate for industrial applications, such as drug transportation, catalysis, and gas adsorption, owing to its thermal, mechanical, and chemical stabilities and diversity in preparation method [72, 73] Generally, to enhance gas adsorption performance, the adsorbent should be meliorated for a relatively high surface area, suitably tuned pore structure, and a large number of adsorption sites having an affinity toward objective gases [73-77] Defect engineering is now considered a promising approach to tune properties of MOFs, including pore structure and specific surface area via the formation of missing linkers and/or clusters, leading to enhancement of gas adsorption capacity Such defects can be generated by surfactant additives or through linker substitution by secondary linkers or modulators [73, 78, 79] By this approach, compared to defect-free UiO-66(Zr), adsorption abilities on defective UiO-66 (Zr) framework were improved for CO2 [73, 80], toluene [81, 82] However, this way often showed a slight improvement in selectivity By another approach, functionalized UiO-66 structure was facilely afforded using terephthalic acid derivatives bearing various groups, such as amino (NH2), hydroxy (OH), carboxylic (COOH), nitro (NO2), methoxy (OCH3), sulfonate (HSO3), and methyl (CH3), during preparation Depending on intrinsic functionalities (chemical property, polarity, and size), the UiO-66-derived materials exhibited active sites that strongly interacted with CO2adsorbate, leading to the improvement of CO2 uptake capacity and selectivity [83, 84] A similar result has been observed on UiO-66-based adsorbent applied for toluene adsorption [54, 58] Recently, a combination of both the defective structure and functionalization on Zr-based UiO-66 was investigated on UiO-66 The obtained materials exhibited impressive results in both adsorption capacity and selectivity in comparison to original UiO-66 material owing to the possession of a defective structure and functional groups having the strong affinity to adsorbates [77, 85] However, the alternative ligands are usually much more expensive than the original ligand (H2BDC), which replaced terephthalic acid (H2BDC) in the preparation of UiO-66 it directly affects production cost In another context, Koutsianos and co-workers [72] successfully introduced nitrogen moieties into the defective structure

of UiO-66 through the post-synthetic defect exchange (PSDE) technique The resulting materials showed a higher CO2 adsorption capacity than that of the original material without

functional groups, depending on the kinds of used N-moieties, owing to a partial functionalization in the defective structure However, this strategy spent two-step in during the synthesis process, leading to an increment of production cost

1.2 Motivation

Contaminated air is becoming more and more serious because of the development of industrial processes and increment in life needs of human Carbon monoxide (CO), carbon dioxide (CO2), and toluene (C7H8) are toxic air pollutants that cause harmful impacts on human health and the ecosystem However, they are widely used as raw materials in many industrial applications such as chemical, petrochemical, pharmaceutical, and cosmetic industry There are many techniques employed for separating gas such as absorption, cryogenic condensation/distillation, adsorption, and membrane Among them, adsorption is more interested than the others because of the requirement low cost and energy-efficient, simplicity in operating process, and high pureness and performance As well-known, the absorbent is a key factor deciding on gas separation performance Metal organic frameworks are known to possess exceptionally large surface area and pore volume, and tunable pore system and surface chemistry, which can be promising candidates for adsorbents as well as support materials

To prepare CO-selective adsorbents, Cu(I) species are incorporated into MOFs support such as MIL-100(Fe), MIL-101(Cr) These MOFs are typically prepared in batch reactors assisted by an electric oven with long reaction times (from 12 h to a few days), high energy consumption, and produce only a small MOF amount Moreover, compared to conventional support, Cu(I)-incorporated MOF adsorbents exhibited high CO adsorption capacity as well as high CO/CO2, CO/N2 However, for practical applications, Cu(I)-based π-complexation adsorbent requires not only high CO adsorption performance but also good air stability due to the possibility of Cu(I) oxidation Therefore, a large amount of support should be successfully fabricated to prepare CO-selective adsorbent with large CO adsorption capacity, high CO selectivity, and excellent air stability, which can be applied on

an industrial scale

For CO2 adsorption, UiO-66(Zr) (UiO = University of Olso) showed relatively modest

CO2 adsorption properties in both uptake capacity and selectivity Although, it possesses exceptional properties such as superior specific surface area, and easily modifiable chemical properties of the surface and their porosity and high stability, simultaneously diversity in synthesis There are many reports exhibiting that CO2 uptake capacity was improved by using defect engineering, but CO2/N2 or CO2/CH4 selectivity was still pretty modesty In order to enhance both adsorption capacity and selectivity, several functional groups as adsorption sites having an affinity toward CO2 were introduced into the UiO-66’s structure

by using an alternative linker Recently, a combination of both the defective structure and functionalization on Zr-based UiO-66 was applied for enhancing CO2 adsorption performance However, these approaches bring a high production cost due to either using an expensive alternative linker or a pathway with multi-step preparation So, a cheap adsorbent with a large CO2 adsorption amount, high CO2/N2 selectivity, and regenerability is intensely desired, contributing to the reduction of CO2 recovery cost

In addition, it was reported that defective framework of Zr-based MOF like UiO-66, UiO-67 via the formation of missing linker and/or clusters, contributing to the improvement

of toluene uptake capacity Similar to CO-selective adsorbent, many defected-framework absorbents were prepared by slow and small-scale solvothermal synthesis which requires a long reaction time (12 h ̴ 24 h) Furthermore, this synthetic approach is not suitable for large-scale production Shortening reaction time could potentially reduce production costs Until now, report on the defect engineering of MOFs within a short reaction time under continuous flow conditions is still limited, although this strategy could be highly promising for the production of surface-modified MOFs Otherwise, to apply the gas adsorption process efficiently in an industrial scale such as pressure swing adsorption (PSA) or vacuum swing adsorption (VSA), the as-prepared MOF powder should be shaped to the body with a bigger size like membranes, pellets, beads owing to convenience in transportation, easy recovery and minimal pressure loss in fixed-bed adsorption system, leading to enhancing adsorption

efficiency So, a defective-UiO adsorbent prepared in a short time, high quality, easy large scale, and shaped in the body is essential to apply on an industrial scale

1.3 Research objectives

In order to meet one of the requirements for application on an industrial scale, the obtained adsorbent has high adsorption capacity and selectivity, stability, generability, potential to large scale, and low production cost Based on the requirements, there are five main objectives in this work

1) To develop several pathways in the synthesis of MIL-100(Fe), 66, and

UiO-67 under assisting microwave heating in both batch and continuous tubular reactor The continuous tubular reactor (diameter - 2 mm and length - 8 m) was employed

to produce MIL-100(Fe) and UiO-67 only with resident time around 50 min and

10 min, respectively Whereas UiO-66 was prepared in a batch reactor (diameter –

100 mm) with a reaction time of 20 min Consequently, the obtained product quality is similar to that of the conventional method with the same recipe in spite

of shortened reaction time significantly This demonstrated that MIL-100(Fe), UiO-66, and UiO-67 have the potential to large scale which meets the requirement

in the air

3) To develop a novel strategy in preparation of CO2-selective UiO-66 adsorbent by combining both the defective structure and functionalization on UiO-66 through using a defective linker attached to several functional groups Herein, 4-amino benzoic acid plays a role as a defective linker The obtained absorbent possesses a