Enhanced Separation Performance for CO2 gas of Mixed-Matrix

Jul 23, 2017 - Tax reform law will boost industry but worries higher ed officials. The most sweeping overhaul of the U.S. tax code in 31 years was sig...
3 downloads 12 Views 6MB Size
Subscriber access provided by UNIV OF WESTERN ONTARIO

Article

Enhanced Separation Performance for CO2 gas of Mixed-Matrix Membranes Incorporated of TiO2/Graphene Oxide: Synergistic Effect of Graphene Oxide and Small TiO2 Particles on Gas Permeability of Membranes Ting Wang, Caihong Yang, Chun-Li Man, Li-guang Wu, WanLei Xue, Jiangnan Shen, Bart Van der Bruggen, and Zhuan Yi Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b02191 • Publication Date (Web): 23 Jul 2017 Downloaded from http://pubs.acs.org on July 24, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

5

Enhanced Separation Performance for CO2 gas of Mixed-Matrix Membranes Incorporated of TiO2/Graphene Oxide: Synergistic Effect of Graphene Oxide and Small TiO2 Particles on Gas Permeability of Membranes

6

Ting Wanga, Cai-hong Yanga, Chun-Li Mana, Li-guang Wua, *, Wan-Lei Xuea, Jiang-nan Shenb,

7

Bart Van der Bruggenc, Zhuan Yia

8

a

School of Environ. Sci. & Eng., Zhejiang Gongshang University, Hangzhou, 310012, China

9

b

Center for Membrane and Water Science, Ocean College, Zhejiang University of Technology,

1 2 3 4

10

Hangzhou 310014, China

11

c

12

KU Leuven, Celestijnenlaan 200F, B-3001 Leuven, Belgium

13

Corresponding Author: Li-guang Wu, e-mail: [email protected].

14

ABSTRACT. This study combines the adsorption layer nanoreactor synthesis (ALRS) method

15

with in situ polymerization to prepare high-performance separation hybrid membranes

16

containing TiO2-graphene oxide (TiO2-GO). TiO2 nanoparticles were initially generated on the

17

GO surface through ALRS. Hybrid membranes containing A-TiO2-GO were then prepared

18

using in situ polymerization. The inhibition of aggregation by the strong combination between

19

the small TiO2 particles and the GO surface led to a homogeneous distribution of TiO2-GO in

Department of Chemical Engineering, Process Engineering for Sustainable Systems (ProcESS),

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 32

20

the polymeric matrix. This, in turn, caused a significant improvement of the membranes

21

performance for CO2 separation. Both the small TiO2 particles and GO played an important role

22

in increasing the gas separation performance of membranes. In particular, TiO2 small particles,

23

with a polar surface, promoted the CO2 diffusion through the membranes, while GO addition

24

significantly increased the solubility selectivity of the membranes towards CO2.

25

INTRODUCTION

26

As a high-efficiency and pollution-free method, polymeric membranes have widely been

27

used in various separation processes for environmental, pharmaceutical, chemical, and many

28

other application fields

29

polymer matrices to prepare novel hybrid membranes attracted much attention due to their

30

desirable organic and inorganic properties

31

membranes are often employed for separation of CO2 or CH4 from gas mixtures

32

researchers have designed and fabricated many novel organic-inorganic polymeric hybrid

33

membranes containing zero-dimensional, one-dimensional, and two-dimensional nanomaterials.

34

Examples include TiO2 and SiO2 nanoparticles

35

nanotubes

36

particularly potential materials due to their excellent CO2 adsorption 14, 15. For instance, Lee et

37

al.

38

CO2/CH4 mixtures. This was due to the single-atom-thick length of graphene, which provides a

39

high selectivity and fast transport properties of gas molecules compared to carbon nanotubes. In

40

most studies on separation gas membranes, permeation of gas molecules through nonporous

41

organic−inorganic membranes is thought to occur via solution and diffusion 17, 18. However, two

16

10, 11

1, 2

. Over the last decades, the addition of various nanomaterials into

3-5

. In particular, organic-inorganic hybrid 6, 7

. Currently,

8, 9

, and graphene/graphene oxide (GO)

, single-walled and multiwalled carbon 12, 13

. Graphene and graphene oxide are

found that graphene membranes efficiently separate CO2 from CO2/O2, CO2/N2, and

ACS Paragon Plus Environment

Page 3 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

42

challenging aspects remain. One challenge is to improve the distribution graphene or GO

43

nanosheets in the polymer matrix. A second challenge concerns the weak influence on both the

44

free volume of membranes and gas diffusion processes after the addition of graphene or GO.

45

In-situ polymerization is an attractive method for fabrication of hybrid membranes containing

46

nanomaterials due to its facile and effective prevention of nanomaterials aggregation during

47

polymerization

48

functional groups of the nanosheets can also improve their distribution in polymer matrices 21, 22.

49

For example, Kim et al.

50

characterization of the resulting films by transmission electron microscopy revealed that

51

dispersion of GO nanosheets in the solvent significantly improved. A previous study showed that

52

GO sheets could be homogeneously distributed in the polyurethane (PU) polymer matrix as more

53

oxygenated functional groups were introduced on the surface of GO sheets 20.

19, 20

. In addition, surface modification of graphene or GO through changes in

23

functionalized graphene nanosheets with polyethylene;

54

On the other hand, studies on the effective enhancement of gas diffusion through hybrid

55

membranes containing graphene or GO are rare. According to the literatures on hybrid

56

membranes containing nanomaterials

57

significantly increases the free volume of polymer membranes, thus enhancing gas diffusion

58

through the hybrid membranes. Thus, it may be interesting to add nanoparticles into hybrid

59

membranes containing graphene or GO for improving gas diffusion. In addition to this, it was

60

found that adding nanoparticles onto the GO surface or into the lamellar structure of GO can

61

effectively

62

nanoparticle–graphene composites with partially exfoliated graphene morphology. Face-to-face

63

aggregation of graphene sheets was prevented by Pt particles of 3–4 nm in size. The

prevent

the

aggregation

8, 9

, the addition of zero-dimensional nanoparticles

of

GO.

Samulski

ACS Paragon Plus Environment

et

al.

24

reported

metal

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

64

Ptnanoparticles acted as spacers in the resulting Pt–graphene composite to produce a

65

mechanically exfoliated and high-surface-area material for supercapacitors and fuel cell

66

applications.

67

This study aims to design hybrid membranes incorporating TiO2-GO with three-dimensional 19, 25, 26

68

networks. Based the literatures and our former works

69

synthesis (ALRS) method was first employed to prepare TiO2-GO. ALRS method is one of the

70

latest developments in microscale reactor technology

71

preparing quantum-sized nanoparticles and controlling nanoparticle microstructures such as

72

particle size

73

nm have been prepared on a SiO2surface and a MWCNT surface. In this study, TiO2-GO

74

samples were first prepared using ALRS and Precipitation Method, respectively. Afterwards,

75

the resulting hybrid membranes incorporated different TiO2-GO samples were fabricated

76

through in situ polymerization for gas separation. The detailed effects of some key factors were

77

studied in combination with gas solution and gas diffusion aspects through the membranes.

78

EXPERIMENTAL

79

Materials

19, 25, 26

. In our previous studies

, Adsorption Layer Reactor

19, 25, 26

. This method is suitable for

19, 25

, well-distributed nanoparticles smaller than 7

80

Graphite powder (G, 8000 mesh) was purchased from Reagent Chemical Manufacturing

81

(Shanghai, China). Sulfuric acid (H2SO4), potassium permanganate (KMnO4), and sodium

82

nitrate (NaNO3) were obtained from Shanghai Reagent Factory (Shanghai, China). Analytical

83

grade tetrabutyl titanate was received from Reagent Chemical Manufacturer (Shanghai, China),

84

and was used without further purification. Analytical grade ethanol purchased from Reagent

ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

85

Chemical Manufacturer (Shanghai, China) was first distilled and then stored in a 4 Å molecular

86

sieves

87

4,4’-diaminodiphenylether

88

N,N-dimethylformamide (DMF) were purchased from Reagent Chemical Manufacturing

89

(Shanghai, China). High-purity (>99.99%) CO2 and N2 gases were purchased from Hangzhou

90

Jingong Gas Co. Ltd.

91

Preparation of TiO2-GO samples

prior

to

use.4,4'-(Hexafluoroisopropylidene) (ODA),

diphthalic

N,N-dimethylacetamide

anhydride (DMAC),

(6FDA), and

92

GO nanosheets with hydrophilic groups were synthesized through the oxidization of graphite

93

powder by improved Hummers method followed by ultrasonication (as shown in Scheme S1 in

94

SI), referring to the preparation of GO-4 in our previous work (The preparation conditions are

95

listed in SI) 20.

96

Preparation of TiO2-GO through ALRS method. The preparation process was reported 19, 25

97

elsewhere

. 0.5 g of as-prepared GO and 200 mL of absolute alcohol were added into a

98

triflask. GO was then well dispersed in alcohol and 1.5 mL or 3.0 mL of water was added into

99

the reaction system under constant stirring at 25 °C. A water-rich adsorption layer was

100

gradually formed on the GO surface because of the selective adsorption capacity of GO. Figure

101

S1 (Supporting Information, SI) gives the adsorption curves of water by GO. Addition of more

102

water from 1.5 mL to 3.0 mL induced more water adsorption into the GO nanosheets, yielding a

103

thick water-rich adsorption layer consistent with previous results. After reaching the adsorption

104

equilibrium (24 h), tetrabutyl titanate dissolved in ethyl alcohol (30 mL, 7.6 g·L−1) was added at

105

a rate of 0.85 mL·min−1.It is worth noting that tetrabutyl titanate could diffuse into the

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19, 25

106

adsorption layer to react first with water

. The reaction was complete after 5 h, and the

107

reaction products were obtained through several centrifugation–redispersion–washing cycles

108

followed by drying at room temperature. The centrifugation condition was 12,000 rpm and

109

centrifuged for 10 minutes each time. The TiO2-GO samples were denoted as A-TiO2-GO1.5 for

110

the sample prepared with 1.5 mL water and A-TiO2-GO3.0 for the sample prepared with 3.0 mL

111

water.

112

Preparation of TiO2-GO by the precipitation method. For comparison, the precipitation

113

method was employed to synthesize TiO2-GO samples. A volume of 200 mL of absolute

114

alcohol and 1.5 mL or 3.0 mL of water was added into a triflask. Tetrabutyl titanate dissolved in

115

ethanol was then added to react at 25 °C and after 4h, 0.5 g of as-prepared GO was added into

116

the reaction system. After completion (5 h), the mixture was centrifuged–dispersed–washed

117

several times to obtain the final product, which was then dried at room temperature. The

118

TiO2-GO samples were denoted as P-TiO2-GO1.5 for the one prepared with 1.5 mL of water and

119

P-TiO2-GO3.0 for the sample prepared with 3.0 mL of water.

120

Fabrication of mixed matrix membrane containing different TiO2-GO samples

121 122

The in-situ polymerization process of PU mixed matrix membranes containing different TiO2-GO samples was referred to our former work 20.

123

The in-situ polymerization of polyimide (PI) mixed matrix membranes containing different

124

TiO2-GO samples: The as-prepared TiO2-GO samples were first dispersed in 12 mL DMAC

125

using an ultrasonic probe (KQ-300TDE; 300 W, 80 kHz) for 40 min. Subsequently, ODA

126

(0.8970 g) was added into the TiO2-GO suspension. After 10 min of constant stirring, 6FDA

ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

127

(2.000 g) was sequentially added to the mixture under vigorous stirring at a temperature below

128

10°C to initiate the polymerization process. The molar ratio of BPDA to ODA was 1.005 to 1.

129

When the reaction viscosity reached approximately 300 mPa·s, the reaction mixture was cast

130

onto a glass plate and dried at room temperature for 30 min. The resulting hybrid membranes

131

with different GO contents were then incubated and subjected to continuous polymerization in a

132

vacuum oven at a heating rate of 2 °C/min at 80 °C for 2 h and at 150 °C, 240 °C and 300 °C for

133

1 h. As shown in Figures S2 and S3 (SI), the high thermal stability of PI polymer materials also

134

ensured the morphology stability of the PI and PI hybrid membrane, after polymerization process

135

as high temperature. The thickness of the polymer membrane top was estimated to about 25 µm,

136

as measured by scanning electron microscopy (SEM).

137

At first, the GO content in all fabricated hybrid membranes was 1.0 wt%, which was optimal

138

for stability and homogeneous distribution in the polymer matrice, and led to the highest

139

separation performance in previous work

140

membranes for further study.

141

Characterization

20

. Then, the GO content changed in PI hybrid

142

The morphologies of the TiO2-GO samples and their PI hybrid membranes were

143

characterized by transmission electron microscopy (TEM, JEM-1230, Jeol Co., Ltd.). The

144

structure of the TiO2-GO samples was determined by Fourier Transform Infrared (FTIR)

145

spectroscopy (Nexus-670, Nicolet Co.). The chemical composition and the state of the elements

146

present on the GO nanosheet surface were investigated with X-ray photoelectron spectroscopy

147

(XPS) measurements using an ESCA-2000, VG Microtech Ltd. Hybrid membranes containing

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 32

148

different TiO2-GO samples were then analyzed using an S-4700 SEM (Hitachi Co., Ltd.,

149

Japan).

150

Gas permeability measurements

151

The

pure

gas

permeability

values

were

determined

using

the

27-30

152

constant-volume/variable-pressure method

. The detailed measure process was indicated in

153

the SI. The determination of two gases permeabilities, the diffusivity, the solubility and the ideal

154

selectivity were also listed in SI.

155

Sorption measurement

156

Sorption was measured with a pressure decay sorption system, referring to the literatures 31, 32.

157

To start, the entire system was degassed overnight. The reservoir cell was filled with a known

158

amount of test gas and equilibrated thermally for 10-15 min. The valve between the reservoir

159

and sample cell was then opened to charge the sample cell with gas, and LabView recorded the

160

pressure decay in both cells. Once the pressure reached equilibrium, the run was stopped and

161

the amount of sorbed gas was calculated based on a mole balance. All sorption isotherms were

162

measured in a pressure range from ~69 kPa (10 psia) to ~1100 kPa (160 psia). The pure gas

163

sorption measurements at 30 °C, 35 °C, 40 °C, and 45 °C on different membranes were

164

examined.

165

RESULTS AND DISCUSSION

166

Morphology of TiO2-GO samples

167

TEM analysis. The TEM images of TiO2-GO samples are shown in Figure 1 and Figure S4.

168

The highly transparent nanosheets with yarn-like sheet structure are characteristic of the GO

ACS Paragon Plus Environment

Page 9 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

169

morphology, which indicative of a monolayer or at most few layers of GO. These findings are

170

consistent with previous work

171

Figure 1 and Figure S4 (SI), shown as black points. The TEM images shown in Figure S4 and

172

Figure 1 also indicate that differences exist in the morphologies of TiO2-GO samples prepared by

173

two synthetic methods.

19, 20

. The morphology of TiO2 particlescan also be observed in

a

b

c

d

174

175 176

Figure 1 TEM images of differentTiO2-GO samples.

177

a. P-TiO2-GO1.5; b. P-TiO2-GO3.0; c. A-TiO2-GO1.5; d. A-TiO2-GO3.0

178

Due to absence of an adsorption layer in precipitation method, there is no restriction on the

179

growth of TiO2 particles. The TiO2 particles are formed and grew freely in the alcohol phase to

180

become large particles. In addition, no significant interaction in two P-TiO2-GO samples is

181

generated, except for some possible physical interactions between the particles and GO

182

nanosheets, like electrostatic adsorption or van der Waals forces. Small TiO2 particle with size of

183

several nanometers are very easy to agglomerate to form a large particles without strong

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

184

protection by GO nanosheets. So, large particles reaching micrometer size are observed in the

185

TEM with wide range of two TiO2-GO samples by precipitation method, as pointed out by the

186

red dotted circles in Figure S4 (SI). And there are also some several significant aggregations of

187

TiO2 particles in two P-TiO2-GO samples, as shown in Figure 1 and Figure S4. The increase in

188

water content in the preparation process from 1.5 mL to 3.0 mL meant more water distributed in

189

the alcohol, so more TiO2 particles were formed in the alcohol, which is noticeable in the

190

significant TiO2 aggregates and large TiO2 particles display in Figure 1b and Figure S4b.

191

However, as the ALRS method uses a water-rich adsorption layer with several nanometers thick

192

on the GO surface as a nanoreactor, this results in smaller size TiO2 nanoparticles of less than 10

193

nm for both TiO2-GO samples prepared by ALRS. Furthermore, the nanoparticles appear more

194

homogeneously distributed on the GO surface, corroborating previous findings 19, 20. When more

195

water is added during the preparation process (3.0 mL), the adsorption layer on GO surface

196

becomes thicker, as shown in Figure S1. Thus, the size of TiO2 particles formed on the GO

197

surface becomesa little larger in Figure 1d compared to Figure 1c.

198

XPS analysis. XPS profiles for Ti2p of different samples are shown in Figure S5 of the

199

Supporting Information. The Ti2p levels of all samples indicate that two peaks at approximately

200

464.6 and 458.3 eV, assigned toTi2p1/2 and Ti2p3/2, respectively33. The data also suggest that Ti

201

existed as Ti4+ in all TiO2-GO samples. The C/O atomic ratio of pristine GO and different

202

TiO2-GO samples has been listed in Table S1 (SI). From the data in the table, the formation of

203

TiO2 particles causes a few decreases in C/O atomic ratio of samples.

204

Figures S6 and S7 display two peaks in the XPS profiles of GO and TiO2-GO samples. For

205

further comparison, the XPS profiles of C1s were deconvoluted. The results are summarized in

ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

206

Table 1, including the oxygen/carbon intensity ratio, the ratio of sp2- to sp3-bonded carbon

207

atoms, as well as the intensity ratio of hydroxyl, carboxyl, and epoxy groups with respect to the

208

C-C peak 34. Table 1 also indicates that the formation of TiO2 particles reduces the number of

209

oxygenated functional groups on the GO surface, due to plausible combinations of TiO2

210

nanoparticles with the GO surface. This reduction also changes carbon atoms from

211

sp3-hybridized orbitals to sp2-hybridized orbitals, consistent with some published reports

212

investigating the effect of nanoparticles

213

induced TiO2 particles in the adsorption layer will combine with the GO surface. Therefore, the

214

number of oxygenated functional groups forms on the GO surface in A-TiO2-GO samples will

215

be smaller than those of P-TiO2-GO samples. In addition, the increase in the number of

216

nanoparticles has a boosting effect. This results in the highest sp2/sp3 ratio of the A-TiO2-GO

217

sample prepared with 3.0 mL water among all samples indicated in Table 1, due to the presence

218

of large amounts of smaller TiO2 particles formed on the GO surface (as shown in the TEM

219

images).

220

Table 1 XPS data of different GO samples.

221

35

. During the preparation process of A-TiO2-GO, all

Sample

sp2/sp3*

C-C/%

C-OH /%

O-C=O/%

C-O-C/%

GO

0.95

25.77

54.83

5.62

13.78

P-TiO2-GO1.5

1.02

29.76

51.79

5.01

13.46

P-TiO2-GO3.0

1.08

31.83

51.06

5.52

11.59

A-TiO2-GO1.5

1.22

43.65

44.34

4.59

7.42

A-TiO2-GO3.0

1.77

51.94

41.02

7.04

0.00

Note: *Ratio of sp2- to sp3-bonded carbon atoms in different samples.

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

222

The formation of TiO2 nanoparticles on the surface of different samples is further confirmed

223

by the results of FTIR and Raman, as shown in Figures S8 and S9 (SI). A comparison of

224

TiO2-GO with GO samples in FTIR spectra shows that the resonance peaks of C-O, C-OH, and

225

absorbed hydroxyl groups in TiO2-GO samples underwent weakening, indicative of a reduction

226

in oxygenated functional groups on the GO surface during the formation of TiO2 particles 36, 37.

227

Figure S8 also indicate that the peaks corresponding to C-O, C-OH, and absorbed hydroxyl

228

groups in the FTIR spectra of two A-TiO2-GO are weaker, compared to those in the FTIR

229

spectra of two P-TiO2-GO. This is related to the formation of small TiO2 particles on GO

230

surface, which facilitates the transfer from sp3-hybridized to sp2-hybridized carbon.

231

Raman spectra shows that the intensity of D-band peaks typically assigned to surface defects 38, 39

232

on GO sheets

becomes stronger, after the formation of the TiO2 nanoparticles. During

233

preparation of TiO2-GO using the ALRS method, all the induced TiO2 particles in the adsorption

234

layer will combine with the GO surface and generate many structural defects on the GO surface.

235

As the number of nanoparticles increases, more structural defects are induced on GO surface.

236

This results in the strongest D-band peak of the A-TiO2-GO3.0 sample prepared by ALRS, among

237

GO and four of the TiO2-GO samples (Figure S9, SI).

238

The TiO2-GO was prepared with the goal of improving GO dispersion in the solvent and thus

239

in the polymer matrix. Figures S10 and S11 (SI) confirm the differences in the distribution of GO

240

or the four TiO2-GO samples in DMAC after 24 h (GO content 10 g·L−1). Two A-TiO2-GO

241

samples reveal a better distribution in DMAC, a typical polar solvent. This improves dispersion

242

of GO in DMAC is mainly related to the preventive effect of GO from aggregation in the solvent

ACS Paragon Plus Environment

Page 12 of 32

Page 13 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

243

by the small TiO2 particles. In addition, the introduction of surface defects increases the surface

244

polarity of GO, which enhances the distribution of GO in DMAC. From Figure S11, the poor

245

distribution in DMAC of two TiO2-GO samples prepared by precipitation method can be

246

observed. This is related to the larger TiO2 particles and TiO2 aggregation in the TiO2-GO

247

samples, as demonstrated by TEM.

248

Morphology of different hybrid membranes

249

The different morphologies of TiO2-GO samples prepared by the two different methods cause

250

differences in morphologies and the performance of the hybrid membranes containing TiO2-GO

251

samples. SEM and TEM were employed to characterize the morphologies of the hybrid

252

membranes, containing different TiO2-GO samples, as shown in Figures 2 and S12 (SI). a

b

c

e

f

253 d

254 255

Figure 2 SEM images of PI or PU hybrid membranes containing different TiO2-GO samples.

256

a. GO/PU; b. A-TiO2-GO3.0/PU; c. P-TiO2-GO3.0/PU; d. GO/PI;

257

e. A-TiO2-GO3.0/PI; f. P-TiO2-GO3.0/PI

258

In A-TiO2-GO samples, TiO2 particles with small size interact tightly with the GO surface.

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

259

During polymerization, these small particles would significantly inhibit the aggregation of GO

260

nanosheets and vice versa. The SEM images of the hybrid membranes shown in Figure 2

261

indicate a homogeneous distribution of A-TiO2-GO in both PU and PI polymeric matrices.

262

Using the precipitation method, TiO2 particles are formed in the alcohol bulk by exerting weak

263

interactions with GO nanosheets. This leads to aggregation ofTiO2 particles in the hybrid

264

membrane during PU or PI polymerization, due to the smaller protective effect of GO in

265

P-TiO2-GO samples. The latter is confirmed by TEM analysis of the hybrid membranes (Figure

266

S12, SI).

267

Performance of membranes containing GO and different TiO2-GO samples

268 269

Figure S13 (SI) and Figure 3 depict the permeability towards two pure gases and CO2/N2 ideal selectivity of PU and PI hybrid membranes incorporated by different TiO2-GO samples. 45 40

PU hybrid membranes PI hybrid membranes

35

CO2/N2 ideal selectivity /α

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

30 25 20 15 10 5 0 a

b

c

d

e

f

270 271

Figure 3 The ideal selectivity of PU or PI and different hybrid membranes.

272

a. PU or PI membranes; b. Membranes with GO; c. Membranes with A-TiO2-GO1.5;

273

d. Membranes with A-TiO2-GO3.0; e. Membranes with P-TiO2-GO1.5; f. Membranes with

274

P-TiO2-GO3.0.

ACS Paragon Plus Environment

Page 14 of 32

Page 15 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

275

The figures show that all hybrid membranes containing TiO2-GO samples have higher CO2

276

permeability and CO2/N2 ideal selectivity than PU or PI membranes. A number of studies 20, 40,

277

41

278

molecules. The strong interactions between GO and CO2 molecules can effectively capture and

279

adsorb CO2 molecules, due to the conjugated π bonds in GO nanosheets, when GO is well

280

distributed in the polymer matrix. The improvement of gas permeability and CO2/N2 ideal

281

selectivity are mainly owing to the addition of TiO2 nanoparticles, compared to those of hybrid

282

membrane containing GO samples. Due to the synergy of GO and small TiO2 particles on the

283

CO2 permeability, the hybrid membranes containing A-TiO2-GO samples have a better

284

performance than other membranes. As the number of TiO2 particles increases through the

285

induction of more water in the preparation process (3.0 mL water), the separation performance

286

towards CO2 of the hybrid membrane incorporated by A-TiO2-GO improves significantly.

suggest GO has a 2D structure with conjugated π bonds, giving strong affinity towards CO2

287

A comparison between PU and PI hybrid membranes reveals that the permeability of PU

288

hybrid membranes is higher than that of PI membranes for pure gases CO2 and N2 under the

289

same conditions. However, the ideal CO2/N2 selectivity of PI hybrid membranes is more

290

pronounced than that of PU hybrid membranes. This might be linked to the different properties

291

of these two polymer materials. To further distinguish between the effects induced by

292

nanoparticles and GO on separation of different hybrid membranes, the diffusivity of CO2 and

293

N2 through different hybrid membranes were first measured (as shown in Figure S14, SI).

294

Based on this, the data of diffusion selectivity (αD) and solubility selectivity (αS) of different PU

295

and PI hybrid membranes are listed in Figure 4, respectively.

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

7

7

αD

αD αS

PU

5

5

Selectivity

6

4

3

4

3

2

2

1

1

0

PI

αS

6

Selectivity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 32

0

a

296

b

c

d

e

f

a

b

c

d

e

f

297

Figure 4 The permeability selectivity of PU or PI and different hybrid membranes

298

a. Polymer; b. Membranes with GO; c. Membranes with A-TiO2-GO1.5; d. Membranes with

299

A-TiO2-GO3.0; e. Membranes with P-TiO2-GO1.5; f. Membranes with P-TiO2-GO3.0.

300

Figure S14 (SI) shows that the addition of GO and TiO2-GO samples increases the diffusivity

301

of CO2 through the membranes, but has a few effects on the diffusivity of N2. Addition of GO

302

and TiO2-GO samples with polar surface will increase both free volume and surface polarity of

303

mixed matrix membrane. Since the difference in electronegativity between C and O in the CO2

304

molecule leads to dipole moments

305

significantly than the diffusion of N2 gas. This result is consistent with the results in our former

306

work

307

diffusion selectivity for CO2 gas than PU or PI membrane, as shown in Figure 4. The increase in

308

both the surface polarity and free volume of mixed matrix membrane can be confirmed by the

309

zeta potential value and water contact angle of membranes, and XRD analysis for different

310

membranes, as shown in Figure S15 (SI) and Figure 5.

311

41

, the CO2 diffusion of all membranes are improved more

20

. Therefore, the hybrid membranes containing GO and TiO2-GO samples have higher

From Figure S15 (SI), addition of GO or TiO2-GO samples into PI matrix will improve the

ACS Paragon Plus Environment

Page 17 of 32

312

surface electrical property and surface polarity of mixed matrix membrane, thus increasing zeta

313

potential values and decreasing water contact angles of hybrid membranes. The hybrid

314

membranes containing two A-TiO2-GO samples have the larger zeta potential values and smaller

315

water contact angles, due to the presence of small TiO2 particles well-distributed on GO surface. f e d c b a

PU

f e d c b a

PI

Intensity /a.u

Intensity /a.u

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

10

12

14

16

18

316

20

22

2θ /°

24

26

28

30

32

10

12

14

16

18

20

22

2θ /°

24

26

28

30

317

Figure 5 XRD patterns of PU or PI and different hybrid membranes.

318

a. PU or PI membranes; b. Membranes with GO;

319

c. Membranes with A-TiO2-GO1.5; d. Membranes with A-TiO2-GO3.0;

320

e. Membranes with P-TiO2-GO1.5; f. Membranes with P-TiO2-GO3.0.

32

321

Figure 5 indicates that all hybrid membranes have a characteristic amorphous peak with no

322

evidence of crystalline reflections. The chain packing can be directly evaluated from X-rays

323

cattering, where the maximum intensity of the amorphous peak is typically related to the

324

intersegmental distance. PU hybrid membranes containing GO and TiO2-GO samples were

325

found to have similar XRD patterns as PU membranes. This suggests that the addition of GO

326

and TiO2-GO induces no significant change in chain packing and free volume of the PU hybrid

327

membrane

328

hybrid membrane in presence of GO and TiO2-GO.

8, 9

. In turn, this results in no significant changes in diffusion selectivities of the PU

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

329

For PI, the change in the characteristic amorphous peaks is obvious, especially in the

330

presence of two A-TiO2-GO samples. After addition of GO, the characteristic amorphous peak

331

becomes wider, indicating an increase in the free volume of the hybrid PI membrane. TiO2

332

particles with larger size did not change the free volume of the hybrid PI membranes, so that the

333

two PI hybrid membranes with incorporated P-TiO2-GO show similar characteristic amorphous

334

peaks as those of GO/PI hybrid membranes. The latter induces similar diffusion selectivities of

335

hybrid PI membranes containing GO and P-TiO2-GO samples. In addition, the characteristic

336

amorphous peaks of the membranes become very wide when two A-TiO2-GO samples are

337

added. It has been reported that polymer chain packing of membrane undergoes a significant

338

disruption, by which the free volume greatly increases due to the presence of smaller TiO2

339

particles. Therefore, hybrid PI membranes incorporated by A-TiO2-GO show elevated diffusion

340

selectivities compared to other hybrid membranes.

341

In addition, the difference in electronegativity between C and O in the CO2 molecule leads to 41

342

dipole moments,

which also improves the diffusion of the CO2 in the hybrid membrane

343

containing GO or TiO2-GO samples with polar surface, thus enhancing the diffusion coefficients

344

of CO2 through the hybrid membranes. Compared to the TiO2-GO samples, addition of GO leads

345

to smaller increase in the diffusion coefficients and selectivity of CO2 gas, even the zeta potential

346

values and water contact angles of membranes with GO sample is similar as that of membranes

347

with TiO2-GO samples. GO is a typical two-dimensional nanomaterial, where nanosheets

348

distributed in the polymeric matrix introduce a few interface gaps and slightly enhance free

349

volume of mixed matrix membrane. Well-distributed TiO2 small particles with polar surface can

350

significantly disrupt the polymer chain packing and increase the free volume of membranes

ACS Paragon Plus Environment

8, 9

.

Page 19 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

351

This, in turn, enhances CO2 diffusion through hybrid membranes. So, the hybrid membranes

352

incorporated by two A-TiO2-GO samples show an elevated diffusion coefficients and diffusion

353

selectivity, since small TiO2 particles in polymeric membranes enhance both the free volume of

354

membranes and the CO2 diffusion through membranes. TiO2 particles with larger size did not

355

promote the CO2 diffusion process. Moreover, the formed aggregates within hybrid membranes

356

containing P-TiO2-GO samples (SEM images) inhibit the diffusion of gases. Therefore, the CO2

357

diffusivity coefficients and selectivity of hybrid membranes containing P-TiO2-GO is similar or

358

slightly lower than that of GO/polymer membranes.

359

From Figure 4, it is also found that the hybrid membranes incorporated by two A-TiO2-GO

360

samples show the similar solubility selectivity as that of membranes with GO, unlike the obvious

361

different effects on the diffusion process. And this is caused by the strong adsorption capability

362

of GO and weak adsorption capability of TiO2 for CO2 gas. In many works

363

interactions between GO and CO2 molecules can effectively capture and adsorb CO2 molecules,

364

due to the conjugated π bonds in GO nanosheets. Well-distributed TiO2 small particles with polar

365

surface can only form physical interactions with the CO2 gas. The significant increase in

366

solubility selectivity for CO2 gas of membranes containing GO and TiO2-GO samples can be

367

confirmed by the CO2 and N2 sorption isotherms of different membranes, as shown in Figure 6,

368

Figures S16 and S17 (SI).

ACS Paragon Plus Environment

40, 41

, the strong

Industrial & Engineering Chemistry Research

110

110

CO2 -30 oC CO2 -35 oC

90

N 2 -40 oC N 2 -45 oC

60

a

50 40 30

90

N2 -30 o C N2 -40 o C

70

N2 -45 o C

b

60 50 40

N2 -35 o C N2 -45 o C

40

20

10

10

1000

0

1200

0

0

200

400

110

110

CO2 -30 o C CO2 -35 o C

100

90

CO2 -45 o C N 2 -30 oC

80

80

C /(cm3(STP)/cm3(pol))

N 2 -35 oC

70

N 2 -40 oC N 2 -45 oC

60

800

1000

0

1200

d

50 40 30

110

70

90 80

60

e

50 40

70

20

10

10

P /kPa

800

1000

1200

1200

f

10

0 600

1000

40 30

400

800

50

20

0

600

60

20

200

400

CO 2 -30 o C CO 2 -35 o C CO 2 -40 o C CO 2 -45 o C N 2 -30 o C N 2 -35 o C N 2 -40 o C N 2 -45 o C

100

30

0

200

P /kPa

CO2 -30 oC CO2 -35 oC CO2 -40 oC CO2 -45 oC N 2 -30 oC N 2 -35 oC N 2 -40 oC N 2 -45 oC

100

CO2 -40 o C

90

600

P /kPa

P /kPa

C /(cm3(STP)/cm3(pol))

369

800

c

50

10

600

N2 -40 o C

60

30

400

N2 -30 o C

70

20

200

CO2 -45 o C

80

N2 -35 o C

20

0

370

CO2 -40 o C

30

0

CO2 -35 o C

100

CO 2 -45 oC

80

N 2 -35 oC

C /(cm3(STP)/cm3(pol))

C /(cm3(STP)/cm3(pol))

90

N 2 -30 oC

70

CO2 -30 o C

CO 2 -35 oC CO 2 -40 oC

CO2 -45 oC

80

110

CO 2 -30 oC

100

CO2 -40 oC

C /(cm3(STP)/cm3(pol))

100

C /(cm3(STP)/cm3(pol))

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 32

0

0

200

400

600

800

1000

1200

0

200

400

P /kPa

600

800

1000

1200

P /kPa

371

Figure 6 CO2 and N2 sorption isotherms of different hybrid membranes at 30 °C, 35 °C, 40 °C,

372

and 45 °C

373

a. PU Membranes with GO; b. PU Membranes with A-TiO2-GO1.5;

374

c. PU Membranes with A-TiO2-GO3.0; d. PI Membranes with GO;

375

e. PI Membranes with A-TiO2-GO1.5; f. PI Membranes with A-TiO2-GO3.0.

376

(The GO content in all hybrid membranes was 1.0 wt%)

377

First, all CO2 and N2 sorption isotherms at the different temperatures along with the Langmuir

378

mode fitting, which is consistent with the results in many literatures 31, 32. The sorption isotherms

379

also show that all membranes have an obviously stronger adsorption capacity for CO2 gas than

380

N2 gas, due to the strong interactions between CO2 molecules and GO in polymer matrix. So, the

381

membranes including two A-TiO2-GO samples have the similar sorption isotherms as those of

382

membrane containing GO sample. This also verifies that addition of TiO2 small particles shows a

383

few effects on the increase in solubility selectivity for CO2 gas of membranes.

384

The difference in CO2 adsorption between GO and TiO2-GO samples is further confirmed by

ACS Paragon Plus Environment

Page 21 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

385

TPD-CO2 results (Figure S18, SI). Figure S18 shows strong intensity peaks ranging from 200 to

386

350 °C in the CO2 desorption curves of GO and TiO2-GO samples, suggesting that all samples

387

strongly absorbed CO2 and thus a strong affinity exists between CO2 and GO or TiO2-GO

388

samples. The P-TiO2-GO shows similar desorption curves as GO, indicating that the large TiO2

389

particles in P-TiO2-GO slightly changes the adsorption of CO2 gas. For both A-TiO2-GO samples,

390

a wide but weak desorption peak ranging from 350 to 600 °C was present in both the desorption

391

curves. This could be caused by a combination of TiO2 small particles and CO2 molecules.

392

Except this, the strong desorption peak displayed between 200 to 350 °C of A-TiO2-GO samples

393

shifted to higher temperatures compared to that of GO and P-TiO2-GO samples. This

394

demonstrates the strong affinity between A-TiO2-GO samples and CO2 molecules, indicating that

395

the small TiO2particles will slightly increase the solution selectivity of the hybrid membranes.

396

Effect of TiO2-GO content on morphology and performance of membranes

397

Except the influence on polymer materials, GO nanosheets will help to keep the good

398

distribution of TiO2 small particles in polymer matrix and vice verse. The content of GO is thus

399

increased in the fabrication process of hybrid PI membranes in order to further investigate the

400

influence of TiO2-GO samples on the morphology and performance of the hybrid membranes 30.

401

The photographs of different hybrid membranes shown in Figures S19 and S20 confirm that

402

small TiO2particles may help in keeping the homogeneous distribution of GO nanosheets in the

403

polymeric matrix. There are no obvious aggregations observed in all hybrid membranes

404

incorporated by different A-TiO2-GO samples; only a few black points are observed in

405

A-TiO2-GO3.0/PI polymer matrices containing 6.0 wt% GO, indicative of aggregations in the

406

polymer matrix (Figure S19). The change in the morphology of the hybrid membranes is further

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

407

confirmed by SEM analysis (Figure 7), as TiO2-GO is added. A

B

C

D

408

409 410

Figure 7 SEM images of PI hybrid membranes containing different A-TiO2-GO samples

411

A. A-TiO2-GO1.5 and 4 wt% GO in hybrid membranes; B. A-TiO2-GO3.0 and 4 wt% GO in

412

hybrid membranes; C. A-TiO2-GO1.5 and 6 wt% GO in hybrid membranes;

413

D. A-TiO2-GO3.0 and 6 wt% GO in hybrid membranes

414

For A-TiO2-GO1.5 samples, no obvious aggregation are observed in the hybrid membrane due

415

to the protection of the small particles on the nanosheets, as well as the few aggregates

416

appearing in the A-TiO2-GO1.5/PI membranes at high loading contents (6 wt% of GO). In

417

hybrid membranes incorporated by A-TiO2-GO3.0 sample, a few aggregates are observed in the

418

polymeric matrix when the GO content is above 4 wt%. The TEM images of the A-TiO2-GO3.0

419

sample display many small TiO2 particles on the GO surface. At higher loading contents (4 wt%

420

or 6 wt% of GO), the presence of large numbers of small particles causes spacing volumes

421

between the particles to reduce. Thus, aggregation of small particles greatly increases during the

422

polymerization process and fabrication of the membranes.

ACS Paragon Plus Environment

Page 22 of 32

Page 23 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

423

The protection of the small TiO2 particles on GO nanosheets is confirmed by changes in the

424

CO2 separation performance of the resulting hybrid membranes. Table 2 shows that the CO2 gas

425

permeability and selectivities of the hybrid membranes increases with the loading content. This

426

is explained by the enhanced distribution of the TiO2 particles and GO in the polymer matrix.

427

More TiO2 particles and GO distribute well in polymeric matrix, both the diffusion selectivity

428

(αD) and solubility selectivity (αS) increase with the loading of A-TiO2-GO samples at GO

429

contents below 4 wt%. The SEM data further reveal that aggregation of particles or nanosheets

430

decreases selectivities of the hybrid membranes containing two A-TiO2-GO samples. Therefore,

431

both the diffusion selectivity (αD) and solubility selectivity (αS) both reduce when GO content

432

in A-TiO2-GO1.5/PI reaches 6.0 wt%. At GO contents above 4 wt%, a few aggregates appear in

433

A-TiO2-GO3.0/PI hybrid membranes, which induce a decline in the membrane performance. The

434

changes in permeability and selectivity of different membranes in pure gas system can be

435

confirmed again by those in mixed gas system, as shown in Table S2 (SI).

436

Table 2 Separation performance of different hybrid membranes.

GO content*

Gas permeability Selectivities /Barrer

Sample /%

CO2

N2

α

αD

αS

1

181.59

5.82

31.20

4.85

6.43

2

211.37

5.37

39.36

5.43

7.25

4

287.56

5.55

51.81

6.38

8.12

6

300.74

6.78

45.36

5.59

7.94

1

211.13

5.28

39.99

6.38

6.27

2

278.26

5.15

54.03

7.26

7.45

A-TiO2-GO1.5/PI

A-TiO2-GO3.0/PI

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

437

Page 24 of 32

4

291.32

5.57

52.30

7.33

7.14

6

301.25

6.68

45.10

6.98

6.46

Notes: *The mass percentage of GO in hybrid membranes.

438

Figure S21 (SI) gives ideal CO2/N2 selectivity values of the membranes (the data in table 2)

439

plotted against CO2 gas permeability. The CO2/N2 selectivity values of the membranes in our

440

work are close to upper bound trade-off line for CO2/N2 separation demonstrated by Robeson in

441

2008 43.

442

CONCLUSIONS

443

Using ALRS as a synthesis method, small TiO2 particles can be generated on the GO surface.

444

The TiO2 particle size formed by ALRS was smaller than that prepared by the precipitation

445

method. This was related to the water-rich adsorption layer of several nanometers formed on the

446

GO surface, which played the role of a nanoreactor in ALRS. TiO2 particles with small size in

447

A-TiO2-GO samples bonded strongly with the GO surface. During in situ polymerization, these

448

small particles significantly inhibit the aggregation of GO nanosheets and vice versa. Therefore,

449

the TiO2-GO samples prepared by ALRS are homogeneously distributed in the hybrid

450

membranes due to the mutual inhibition of aggregation of both small particles and GO

451

nanosheets. This, in turn, significantly increases the CO2 gas permeability and CO2/N2 ideal

452

selectivity. The small TiO2 particles and GO play different roles in increasing the gas separation

453

performance of the hybrid membranes. The small TiO2 particles with polar surface in polymeric

454

membranes enhance the CO2 diffusion through the hybrid membranes. The addition of GO

455

plays an important factor in the significant increase of the solubility selectivity of the

456

membranes due to the strong adsorption of CO2 gas. On the other hand, inhibition of GO

ACS Paragon Plus Environment

Page 25 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

457

aggregation by the small TiO2 particles generates a homogenous distribution of GO in the

458

polymeric matrix up to maximum 4 wt% content, which is four times higher than the GO added

459

content reported in previous work.

460

SUPPORTING INFORMATION

461

Dependences of water adsorbed by GO on adsorption time, Digital photos of PI polymer

462

membrane with or without heating at 350 oC for 1 h, TGA thermograms of PI and different PI

463

hybrid membranes, TEM images with wide range of different TiO2-GO samples, XPS profiles

464

of Ti2p in different TiO2-GO samples, XPS profiles of C1s in GO samples, XPS profiles of C1s

465

in different TiO2-GO samples, FTIR spectra of GO and different TiO2-GO samples, Raman

466

spectra of GO and different TiO2-GO samples, Digital photos of GO in DMAC after 24 h,

467

Digital photos of different TiO2-GO samples in DMAC after 24 h, TEM images of hybrid

468

membranes containing different TiO2-GO samples, The gas permeability of PU or PI and

469

different hybrid membranes, The diffusivity through PU or PI and different hybrid membranes,

470

Zeta potential values of PU or PI and different hybrid membranes, CO2 and N2 sorption

471

isotherms of PU and PI membranes at 30 °C, 35 °C, 40 °C, and 45 °C, CO2 and N2 sorption

472

isotherms of different hybrid membranes at 30 °C, 35 °C, 40 °C, and 45 °C,CO2-TPD results of

473

GO and different TiO2-GO samples, Photographs of PI membrane and GO/PI hybrid

474

membranes, Photographs of PI hybrid membranes containing different A-TiO2-GO samples,

475

and CO2/N2 Ideal selectivities of different PI hybrid membranes were listed in Figures S1-S21.

476

The formation mechanism of GO nanosheets by improved Hummers method was listed in

477

Scheme S1. The C/O atomic ratio in GO and different TiO2-GO samples and Separation

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

478

performance of different hybrid membranes using a mixture of CO2/N2 (1:9 v/v) as test gas

479

were listed in Table S1 and S2. This material is available free of charge via the Internet at

480

http://pubs.acs.org.

481

AUTHOR INFORMATION

482 483

Corresponding Author Phone: +86 571 28008204; Fax: +86 571 28008215 e-mail: [email protected].

484

Notes

485

The authors declare no competing financial interest.

486

ACKNOWLEDGMENT

487

Financial support from the National Natural Science Foundation of China (Contract

488

21376218), the Natural Science Foundation of Zhejiang Province (Contract LY14B060001) and

489

Open Research Fund Program of Collaborative Innovation Center of Membrane Separation and

490

Water Treatment of Zhejiang Province (Contract 2016YB08) are gratefully acknowledged.

491

REFERENCE

492

(1) Fihri, A.; Mahfouz, R.; Shahrani, A.; Taie, I.; Alabedi, G. Pervaporative Desulfurization of

493

Gasoline: A Review. Chem. Eng. Process. 2016, 107, 94–105.

494

(2) Hejna, A.; Kosmela, P.; Formela, K.; Piszczyk, Ł.; Haponiuk, J. T. Potential Applications of

495

Crude Glycerol in Polymer Technology–Current State and Perspectives. Renew. Sust. Energ. Rev.

496

2016, 66, 449–475.

497

(3) Goh, P. S.; Ng, B. C.; Lau, W. J.; Ismail, A. F. Inorganic Nanomaterials in Polymeric

ACS Paragon Plus Environment

Page 26 of 32

Page 27 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

498

Ultrafiltration Membranes for Water Treatment. Sep. Purif. Rev. 2015, 44, 216–249.

499

(4) Nasir, R.; Mukhtar, H.; Man, Z.; Mohshim, D. F. Material Advancements in Fabrication of

500

Mixed-Matrix Membranes. Chem. Eng. Tech. 2013, 36, 717–727.

501

(5) Ng, L. Y.; Mohammad, A.W.; Leo, C. P.; Hilal, N. Polymeric Membranes Incorporated with

502

Metal/Metal Oxide Nanoparticles: A Comprehensive Review. Desalination 2013, 308, 15–33.

503

(6) Wang, S. F.; Li, X. Q.; Wu, H.; Tian, Z. Z.; Xin, Q. P.; He, G. W.; Peng, D. D.; Chen, S. L.;

504

Yin, Y.; Jiang, Z. Y.; Guiver, M. D. Advances in High Permeability Polymer-Based Membrane

505

Materials for CO2 Separations. Energ. Environ. Sci. 2016, 9, 1863–1890.

506

(7) Rezakazemi, M.; Amooghin, A. E.; Montazer-Rahmati, M. M.; Ismail, A. F.; Matsuura, T.

507

State-of-the-Art Membrane Based CO2 Separation Using Mixed Matrix Membranes (MMMs):

508

An Overview on Current Status and Future Directions. Prog. Polym. Sci. 2014, 39, 817–861.

509

(8) Boffa, V.; Parmeggiani, L.; Nielsen, A. H.; Magnacca, G. Hydrophilicity and Surface

510

Heterogeneity of TiO2-Doped Silica Materials for Membrane Applications. Micropor. Mesopor.

511

Mat. 2016, 221, 81−90.

512

(9) Swaidan, R.; Ghanem, B.; Litwiller, E.; Pinnau, I. Effects of Hydroxyl-Functionalization and

513

Sub-Tg Thermal Annealing on High Pressure Pure- and Mixed-Gas CO2/CH4 Separation by

514

Polyimide Membranes Based on 6FDA and Triptycene-Containing Dianhydrides. J. Membr. Sci.

515

2015, 475, 571−581.

516

(10) Hinds, B. J.; Chopra, N.; Rantell, T.; Andrews, R.; Gavalas, V.; Bachas, L. G. Aligned

517

Multiwalled Carbon Nanotubes Membranes. Science 2004, 303, 62–65.

518

(11) Zhang, H. Y.; Guo, R. L.; Hou, J. P.; Wei, Z.; Li X. Q. Mixed-Matrix Membranes Containing

519

Carbon Nanotubes Composite with Hydrogel for Efficient CO2 Separation. ACS Appl. Mater.

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

520

Interfaces, 2016, 8, 29044–29051.

521

(12) Yoon, H. W.; Cho, Y. H.; Park, H. B. Graphene-Based Membranes: Status and Prospects.

522

Phil. Trans. R. Soc. A 2016, 374, 20150024.

523

(13) Shen, L.; Xiong, S.; Wang, Y. Graphene Oxide Incorporated Thin-Film Composite

524

Membranes for Forward Osmosis Applications. Chem. Eng. Sci. 2016, 143, 194−205.

525

(14) Khan, M. M.; Filiz, V.; Bengtson, G.; Shishatskiy, S.; Rahman, M.; Abetz, V. Functionalized

526

Carbon Nanotubes Mixed Matrix Membranes of Polymers of Intrinsic Microporosity for Gas

527

Separation. Nano. Res. Lett. 2012, 7, 504−515.

528

(15) Bonaccorso, F.; Colombo, L.; Yu, G. H.; Stoller, M.; Tozzini, V.; Ferrari, A. C.; Ruoff, R. S.;

529

Pellegrini, V. Graphene, Related Two-Dimensional Crystals, and Hybrid Systems for Energy

530

Conversion and Storage. Science 2015, 347, 1246501−1246509.

531

(16) Lee, J.; Aluru, N. R. Water-Solubility-Driven Separation of Gases Using Graphene

532

Membrane. J. Membr. Sci. 2012, 428, 546−553.

533

(17) Baker, R. W.; Low, B. T. Gas Separation Membrane Materials: A Perspective.

534

Macromolecules, 2014, 47, 6999–7013.

535

(18) Tong, W. S.; Zhang, Y. H.; Zhang, Q.; Luan, X. L.; Duan, Y.; Pan, S. F.; Lv, F. Z.; An, Q.

536

Achieving

537

Oxide/Polymer Composite by Covalent Modification of Graphene Oxide Surface. Carbon 2015,

538

94, 590−598.

539

(19) Wang, T.; Jiang, Y. Y.; Shen, J. N.; Wu, L. G.; Bruggen, B. V.; Dong, C. Y. Preparation of Ag

540

Nanoparticles on MWCNT Surface via Adsorption Layer Reactor Synthesis and Its

541

Enhancement on the Performance of Resultant Polyurethane Hybrid Membranes. Ind. Eng.

Significantly

Enhanced

Dielectric

Performance

ACS Paragon Plus Environment

of

Reduced

Graphene

Page 29 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

542

Chem. Res. 2016, 55, 1043−1052.

543

(20) Wang, T.; Zhao, L.; Shen, J. N.; Wu, L. G.; Bruggen, B. V. Enhanced Performance of

544

Polyurethane Hybrid Membranes for CO2 Separation by Incorporating Graphene Oxide: The

545

Relationship between Membrane Performance and Morphology of Graphene Oxide. Environ.

546

Sci. Technol.2015, 49, 8004−8011.

547

(21 )Suzuki, T.; Yamada, Y.

548

branched Polyimide–Silica Hybrid Membranes. J. Membr. Sci. 2012, 417–418, 193–200.

549

(22) Zhang, Q. P.; Li, Q. L.; Xiang, S. D.; Wang, Y.; Wang, C. Y.; Jiang, W.; Zhou, H.; Yang, Y.

550

W.; Tang, J .Covalent Modification of Graphene Oxide with Polynorbornene by

551

Surface-Initiated Ring-Opening Metathesis Polymerization. Polymer 2014, 55, 6044−6050.

552

(23) Kim, H.; Kobayashi, S.; AbdurRahim, M. A.; Zhang, M. J.; Khusainova, A.; Hillmyer, M.

553

A.; Abdala, A. A. ; Macosko, C. W. Graphene/Polyethylene Nanocomposites: Effect of

554

Polyethylene Functionalization and Blending Methods. Polymer 2011, 52, 1837−1846.

555

(24) Si, Y. C.; Samulski, E. T. Exfoliated Graphene Separated Platinum Nanoparticles. Chem.

556

Mater. 2008, 20, 6792−6797.

557

(25) Wang, T.; Jiang, X.; Wang, J. Adsorption Phase Synthesis: Preparation of Nanoparticles and

558

the Effects of Reactant Distribution. J. Colloid Interface Sci. 2010, 350, 69−74.

559

(26) Beck, A.; Horváth, A.; Szucs, A.; Schay, Z.; Horváth, Z. E.; Zsoldos, Z.; Dékány, I.; Guczi,

560

L. Pd Nanoparticles Prepared by Controlled Colloidal Synthesis in Solid/Liquid Interfacial Layer

561

on Silica. I. Particle Size Regulation by Reduction Time. Catal. Lett. 2000, 65, 33−42.

562

(27) Cong, H. L.; Yu, B. Aminosilane Cross-Linked PEG/PEPEG/PPEPG Membranes for

563

CO2/N2 and CO2/H2 Separation. Ind. Eng. Chem. Res. 2010, 49, 9363−9369.

Effect of Thermal Treatmenton Gas Transport Properties of Hyper

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Yu,

B.;

Cong,

H.

L.;

Zhao,

X.

S.

Hybrid

564

(28)

565

Poly(2,6-diphenyl-1,4-phenyleneoxide) and SiO2 Nanocomposite Membranes for CO2/N2

566

Separation. Prog. Nat. Sci. 2012, 22, 661−667.

567

(29) Car, A.; Stropnik, C.; Yave, W.; Peinemann, K. V. PEG Modified Poly(amide-b-ethylene

568

oxide) Membranes for CO2 Separation. J. Membr. Sci. 2008, 307, 88−95.

569

(30) Wijmans, J. G.; Baker, R. W. The Solution–Diffusion Model: A Review. J. Membr. Sci. 1995,

570

107, 1−21.

571

(31) Fu, S. L.; Sanders, E. S.; Kulkarni, S. S.; Wenz, G. B.; Koros, W. J. Temperature

572

Dependence of Gas Transport and Sorption in Carbon Molecular Sieve Membranes Derived

573

from Four 6FDA Based Polyimides: Entropic Selectivity Evaluation. Carbon 2015, 95,

574

995–1006.

575

(32) Agarwal, R. A.; Gupta, N. K. CO2 Sorption Behavior of Imidazole, Benzimidazole and

576

Benzoic Acid Based Coordination Polymers. Coordin. Chem. Rev. 2017, 332, 100–121.

577

(33) Kong, L.; Wang, C. H.; Zheng, H.; Zhang, X. T.; Liu, Y. C. Defect-Induced Yellow Color in

578

Nb-Doped TiO2 and Its Impact on Visible-Light Photocatalysis. J. Phys. Chem. C 2015, 119,

579

16623−16632.

580

(34) Lee, D.W.; Seo, J.W. Sp2/sp3 Carbon Ratio in Graphite Oxide with Different Preparation

581

Times. J. Phys. Chem. C 2011, 115, 2705–2708.

582

(35) Kaspar, T. C.; Droubay, T.; Chambers, S. A.; Bagus, P. S. Spectroscopic Evidence for Ag(III)

583

in Highly Oxidized Silver Films by X-ray Photoelectron Spectroscopy. J.Phys. Chem. C 2010,

584

114, 21562–21571.

585

(36) Xu, C.; Wang, X.; Zhu, J. W. Graphene–Metal Particle Nanocomposites.J. Phys. Chem. C

ACS Paragon Plus Environment

Brominated

Page 30 of 32

Sulfonated

Page 31 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

586

2008, 112, 19841–19845.

587

(37) Goncalves, G.; Marques, P. A. A. P.; Granadeiro, C. M.; Nogueira, H. I. S.; Singh, M. K.;

588

Grácio, J. Surface Modification of Graphene Nanosheets with Gold Nanoparticles: The Role of

589

Oxygen Moieties at Graphene Surface on Gold Nucleation and Growth. Chem. Mater. 2009, 21,

590

4796–4802.

591

(38) Ferrari, A. C.; Robertson, J. Interpretation of Raman Spectra of Disordered and Amorphous

592

Carbon. J. Phys. Rev. B 2000, 61, 14095−14107.

593

(39) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.; Ferry,

594

J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A.

595

R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G.

596

L. Frontiers, Opportunities, and Challenges in Biochemical and Chemical Catalysis of CO2

597

Fixation. Chem. Rev. 2013, 113, 6621−6658.

598

(40) Sarfraz, M.; Ba-Shammakh M. Synergistic Effect of Adding Graphene Oxide and ZIF-301

599

to Polysulfone to Develop High Performance Mixed Matrix Membranes for Selective Carbon

600

Dioxide Separation from Post Combustion Flue Gas. J. Membr. Sci. 2016, 514, 35–43.

601

(41) Chowdhury, S.; Parshetti, G. K.; Balasubramanian R. Post-Combustion CO2 Capture Using

602

Mesoporous TiO2/Graphene Oxide Nanocomposites. Chem. Eng. J. 2015, 263, 374–384.

603

(42) Stankovich, S.; Piner, R. D.; Nguyen, S. T.; Ruoff, R. S. Synthesis and Exfoliation of

604

Isocyanate-Treated Graphene Oxide Nanoplatelets. Carbon 2006, 44, 3342−3347.

605

(43) Robeson, L. M. The upper bound revisited. J. Membr. Sci. 2008, 320, 390–400.

606

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

607

Table of Contents Graphic

608 609

ACS Paragon Plus Environment

Page 32 of 32