Supporting Information

Selectively Tune Gas Permeation through Ionic Liquid Filled Nanoslits by Electric Field

Wen Ying,† Ke Zhou,† Quangang Hou,† Danke Chen, Yi Guo, Jun Zhang, Youguo Yan,* Zhiping Xu* and Xinsheng Peng*

Electronic Supplementary Material (ESI) for Journal of Materials Chemistry A.This journal is © The Royal Society of Chemistry 2019

Experimental Section

Synthesis of GO Nanosheets: GO nanosheets were prepared by the modified Hummer's

method.1 0.5619 g graphite power, 1.875 g P2O5 and 1.875 g K2S2O8 were added into

9.9 ml concentrated H2SO4, and stirred under 80 °C for 4.5 h. Afterward, the mixture

was diluted into 750 ml deionized water (DIW) and collected by vacuum filtration.

Then the product was washed by DIW till neutral, followed by drying under room

temperature for 24 h. The pre-processed graphite was stirred with 90 ml H2SO4 in ice-

water bath, and 11.25 g K2MnO4 was added slowly. When the mixture dispersed evenly,

it was stirred under 35 °C for 2 h. Next, it was diluted with 187.5 ml DIW in ice-water

bath and stirred for another 2 h under 35 °C. After that, 525 ml DIW and 15 ml H2O2

(30%) were added, and 30 ml HCl (37%) was added 5 m later. The solution was laid

aside. After 2 days, the sediment was collected and washed by centrifuged with HCl

(10%) and DIW by turns for several times. Finally, the product was dried under 50 °C.

Synthesis of GO Membrane and GO-SILM: The GO aqueous dispersion was prepared

by sonicating 0.2 g GO in 1 L DIW. The GO membrane was obtained by filtrating GO

aqueous on an anodic aluminum oxide (AAO) or PC substrate. The GO-supported ionic

liquid membrane (GO-SILM) was fabricated by dropping [BMIM][BF4] or

[EMIM][BF4] on a GO membrane and the membrane was laid aside for 2 h before

spinning with 1000 r/m for 6 s and 7000 r/m for 60 s.

Characterization: The scanning electronic microscopy (SEM) images were taken by a

Hitachi S-4800 field-emission SEM equipped with an energy dispersive X-ray

spectroscopy (EDS) spectrometer and the X-ray diffraction (XRD) patterns were

obtained by SHIMADZU XRD-6000 with a Cu Kα radiation source. Differential

scanning calorimetry (DSC) results were obtained by TA Q200 Differential Scanning

Calorimeter. Atomic force microscopy (AFM) imagines were taken by Bruker

Dimension Edge and the Fourier-transform infrared spectroscopy (FTIR) spectrum was

obtained by Bruker Tensor 27. The C/O values were measured by Kratos, AXIS Supra,

X-ray photoelectron spectrometer.

Gas Separation: Bubble flow meter with total volume of 1 ml with 0.01 ml accuracy

was used to measure gas permeance. The GO-SILM was assembled in an in-line

stainless steel filter holder with effective membrane surface area was 2.2 cm2,

purchased from Millipore. The gas permeance measurements were performed ten times

for each membrane and three membranes were evaluated. Gas permeance (Q) was

calculated according to the equation (1):2

(1)𝑄 =

1𝑃𝑢 ‒ 𝑃𝑑

∗273.15

273.15 + 𝑇∗

𝑃𝑎𝑡𝑚

101.325∗

1𝐴

∗𝑑𝑉𝑑𝑡

where Pu is the upstream pressure, Pd is the downstream pressure (atmosphere in our

testing environment), Patm is the atmospheric pressure, A is the membrane effective

area, T is the temperature, dV/dt is the volumetric displacement rate in the bubble flow

meter. In this work, Pu is 0.16 MPa, A is 2.2 cm2 and T is 273.15 K. Tested gases are

single pure gas. The 50% porosity of the AAO was considered. The ideal selective

factor of two gases is the ratio of two gas permeance:2

(2) 𝛼 =

𝑄1

𝑄2

where Q1 and Q2 are the two permeance, respectively.

The gas permeability was calculated by multiplying permeance (Q) and thickness.

MD Simulations Setup: Classical MD simulations were performed using the large-scale

atomic/molecular massively parallel simulator (LAMMPS).3 The all-atom optimized

potentials for liquid simulations (OPLS-AA) were used for GO,4 which are able to

capture essential many-body terms in interatomic interactions. The AMBER force field

of Cornell and co-workers5 are used for ionic liquid [BMIM][BF4], following functional

form for the potential energy:

𝑉

= ∑𝑏𝑜𝑛𝑑𝑠

𝐾𝑟(𝑟 ‒ 𝑟0)2 + ∑𝑎𝑛𝑔𝑙𝑒𝑠

𝐾𝜃(𝜃 ‒ 𝜃0)2 + ∑𝑑𝑖ℎ𝑒𝑑𝑟𝑎𝑙𝑠

𝐾𝜒[1 + 𝑐𝑜𝑠(𝑛𝜒 ‒ 𝛿)]2 + ∑𝑖 < 𝑗

4𝜀𝑖𝑗[(𝜎𝑖𝑗

𝑟𝑖𝑗)12 ‒ (𝜎𝑖𝑗

𝑟𝑖𝑗)6] + ∑

𝑖 < 𝑗

𝑞𝑖𝑞𝑗/𝑟𝑖𝑗

(3)

In our work, the parameters for AMBER formation parameters developed by Jones de

Andrade and co-workers are employed,6 which can successfully capture the statics and

dynamics properties of [BMIM][BF4]. The van der Waals parameters for interactions

between different atom types are obtained from the Lorentz-Berthelot mixing rule. The

time step is 1 fs. The Ewald summation7 was used in the Cornell force field for the

long-range electrostatic interactions. The SHAKE algorithm was applied for hydrogen

atoms to reduce high-frequency vibrations that require shorter time steps. The force

field parameters for gases (CO2, N2, H2, and CH4) are taken from our previous work.8

Simulations were carried out at 300 K using the Nosé-Hoover thermostat with a

damping time constant of 1 ps.

Modelling Confined ILs: We simulated 300 pairs of [BMIM][BF4] molecules in a

simulation box of 6.3×5.8×10 nm, corresponding 4 layers of ILs (Fig. S9). Two walls

of GO sheets separated by 2.2 nm are used in the z direction to confine the ILs under

EEFs, which, however, are permeable for the gas molecules by turning off the gas-wall

interaction in simulations. Periodic boundary conditions (PBCs) are applied in all

directions, and the box size of 10 nm in the z direction is verified to be large enough to

exclude interaction between periodic images in that dimension. The system was relaxed

with a relatively low pressure of the confined IL (|pz| < 100 MPa, |px| < 50 MPa, and

|py| < 50 MPa). The confining graphene oxide walls are modelled using OPLS-AA

parameters.47 EEFs ranging from -2 V/Å to 2 V/Å are applied in the z direction.

The density probability distribution function (D-PDF) were calculated for the number

of cations and anions, based on their centres of masses in the equilibrium structures

(Fig. 2) The results suggest that, in the absence of EEFs, the IL is layered in channel.2

Under a finite EEF, the D-PDFs of cations and anions are shifted in opposite directions

along the EEF, and the amplitude of shifting increases with the EEF strength. The

asymmetric structures of confined ILs were equilibrated by coupling to a Nosé-Hoover

thermostat for 2 ns before gas adsorption free energy calculations following the method

proposed in the literature (Fig.S10).9,10

Gas Absorption Free Energy Calculations: The gas absorption/desorption free energy

is calculated from the potential of means force (PMF) for gas absorption from vacuum

to surface of confined ILs by performing steered molecular dynamics (SMD)

simulations in the constant-speed mode. The pulling velocity and spring constant are 5

Å/ns and 5 kcal/(mol Å2), respectively. The work of spring force W is recorded as a

function of pulling distance in the z direction. The change in gas adsorption free energy

is calculated from W by using the second-order cumulant expansion formula of

Jarzynski’s equality, ,11 where <…> indicates 𝐺 =< 𝑊 >‒ ( < 𝑊2 >‒ < 𝑊 > 2)/2𝑘𝐵𝑇

ensemble averaging performed over a set of 5 independent MD runs. The representative

PMF curves (G) is illustrated in Fig. S9 and the forward/backward process represents

gas absorption/desorption from the IL.9,10 Here, the difference between the minimum

in PMF related to the vacuum value is defined as the absorption free energy Gabs, which

is 2.37 kcal/mol for CO2 in the absence of EEF, in consistency with the results reported

for CO2 absorbed to the surface of bulk [BMIM][BF4].9 The absorption free energy

Gabs(E) is shown in Fig. S10 as a function of EEF, which is normalized by its value of

in the absence of EEF. As for the desorption process on the permeate side, which is the

reverse process of absorption, the desorption free energy is . We can 𝐺𝑑𝑒𝑠(𝐸) = 𝐺𝑎𝑏𝑠( ‒ 𝐸)

then estimate the permeability as a function of E, , where 𝑝(𝐸) = ~𝑒𝑥𝑝(∆𝐺(𝐸)/𝑘𝐵𝑇)

. The values of ∆G(E) is summarized in ∆𝐺(𝐸) = 𝐺𝑎𝑏𝑠(𝐸) ‒ 𝐺𝑑𝑒𝑠(𝐸) = 𝐺𝑎𝑏𝑠(𝐸) ‒ 𝐺𝑎𝑏𝑠( ‒ 𝐸)

Fig. 3a for all the gas. The results are normalized by the value of in the absence of EEF,

which can be used to quantize the enhancement of P comparing to without EEF.

Free Volume and Anion-cation Interaction Energy Calculations: 190 pairs of ionic

liquid are filled in a 2D membrane packaged by two parallel GO walls with the size of

10.0×5.0×2.2 nm3, as shown in Fig. S11. PBCs were used in all three directions. The

length of box in electric field direction is 10 nm to eliminate the interaction of periodic

images. The constant external electric field are applied in z direction from -2 V/Å to 2

V/Å. For each system, the simulation was conducted in a canonical ensemble (NVT) at

298 K controlled by the Nosé-Hoover thermostat. The free volume is calculated by the

Connolly Surface.12 All the data were calculated after an equilibrium molecular

dynamics of 2 ns. The interlayer spacing of GO sheets is set as 2.2 nm, which is

consistent with the XRD experiments. In the simulations, all atoms in the GO sheets

were fixed except the hydroxyl and epoxy groups.

Figures and tables

Fig. S1 AFM images of GO nanosheets.

Fig. S2 XRD patterns of a typical GO membrane and GO-SILM.

Fig. S3 The SEM imagines of GO-SILMs with different thickness. (a)-(d), the cross-section of GO-SILMs with the thickness of 72 nm, 170 nm, 280 nm, 380 nm, including 0.01, 0.025, 0.04, 0.05 mg GO, respectively.

Fig. S4 The FTIR spectrum of GO, [BMIM][BF4] and GO-SILM.

Fig. S5 The DSC curves of GO, [BMIM][BF4] and GO-SILM.

Fig. S6 The gas permeance and separation of GO-SILMs with different thickness.

Fig. S7 The separation performance (selectivity vs. permeance) of GO-SILMs under EEF compared with other membranes. (a) CO2/H2, (b) CO2/CH4 and (c) CO2/N2. The detail data is shown in Table S3 and from the Supplementary references.

Fig. S8 (a) is the CO2/H2 selectivity of GO-SILM on PC substrate under different EEF; (b) is the CO2/H2 selectivity of GO-SILM with [EMIM][BF4] under different EEF.

Fig. S9 Illustration of molecular simulation for IL film models. The two grey rectangles are GO walls with the position of upper one at z = 0.5 nm. The interaction between gas and GO walls is turned off so gas molecules can permeate into the IL. The gas molecules are pulled by a harmonic spring with its end point moving from z = 0 to z = 1.2 nm in the SMD simulations. The representative PMF curves denote the absorption/desorption processes of gases into/from ILs. The difference between the free energy minimum in PMF and the value in vacuum is defined as the surficial absorption free energy .∆𝐺

Fig. S10 The adsorption free energy (Gabs) calculated for gases in confined IL films under a finite EEF. The results are normalized by the value of in the absence of EEF.

Fig. S11 MD models to study the free volume and anion-cation interaction energy. White, grey, red, blue, purple and green balls denote the hydrogen, carbon, oxygen, nitrogen, fluorine and boron atom respectively.

Table S1 The list of GO weight content in GO-SILMs with different thickness.

Membrane Thickness (nm) Original GO (nm) Weight of GO (mg)

GO-SILM-1 72 30 0.01GO-SILM-2 170 80 0.025GO-SILM-3 280 120 0.04GO-SILM-4 380 160 0.05

Table S2 FTIR spectrum peak shifts of GO-SILMs compared with [BMIM][BF4].

Peak of [BMIM][BF4] (cm-1) Peak of GO-SILM (cm-1) Shift (cm-1) Assignment3122.24 3118.38 -3.9 νC2-H1384.66 1386.59 1.93 wCH2

1373.09 1363.45 -9.6 νC2N1C5, wCH21338.38 1340.31 1.93 νN-Bu, N-Me1284.38 1288.24 3.86 νBF4

1253.52 1247.93 -5.6 tCH2, rC2-H1168.67 1157.10 -12 νN-Bu, N-Me1114.67 1097.31 -17 rCH3rCH3(Bu)1031.74 1037.53 5.79 νBF1014.39 1010.53 -3.9 νBF4

846.61 825.40 -21 rC4-H,rC5-H808.04 811.90 3.86 νC-C-C752.11 754.04 1.93 rCH2

696.19 705.83 9.64 νN-Bu, N-Me651.83 653.76 1.93 νN-Bu, N-Me

Table S3 The original data for Fig. S7.Membrane PCO2(GPU) CO2/H2 CO2/CH4 CO2/N2 Ref

ZIF-90/6FDADAM 12 28 13PSF 19.86 23.12 14DDR 35.8 98 15

MIL-53-NH2/organosilica 430 23.2 16MMP-3/mPSf 3000 78 17

PAMAM dendrimer 61 230 18DNMDAm–DGBAmE–

TMC/PDMS/PS1601 138 19

PVAm–PIP/PS 6500 277 20DAmBS–DGBAmE–

TMC/PDMS/PS5831 0.86 86 21

SIPN:PEGMEA:PEGDA:PEGDME 11.9-14.9 14.7 45.7 22SIPN:PEGDA:PEGDME 7.1-8.8 14.5 65.9 22

Matrimid® ZIF-8 180 41.5 23Matrimid® CuBDC 0.059 88.2 24

6FDA-DAT/Ni2(dodbc) 0.91~1.6 52 256FDA-ODA UiO-66 1.26~2.52 46.1 26

Pebax/[emim][BF4]/GO 981 44 27Pebax/GO 20 13 26 91 28

Pebax/[emim][BF4] 306 15 36 29PAN/[emim][Ac]GO/PTMSP 37 16 39 130 30

Polyactive/GO 2.38~2.86 12 21 73 31GO 110 6 10 20 32

0.1water-[BMIM][BF4]-PES 13.8 60 33[BMIM][BF4]/AgO 14.1 - 28.2 34[BMIM][BF4]/CuO 52.4 - 21 35

Pebax/0.5%Ag/50%[BMIM][BF4] 3.2 61 187.5 36[BMIM][BF4]/LiBF4 13.36 8.25 8.4 37

[BMIM][BF4]/cyanuric chloride 19.2 10.7 11 38GO-[BMIM][BF4] (1050 nm) 68.5 23.7 234 382 2

GO-[BMIM][BF4] (380 nm) without EEF

117 15.3 52.9 86.5 This work

GO-[BMIM][BF4] (380 nm) under positive EEF

259 42.8 262.4 528.8 This work

Table S4 The C and O atomic proportions in GO before and after applying the EEF which were measured by XPS.

Element GO GO +EEFC 2.49 2.57O 1 1

Table S5 Lennard-Jones parameters and atom charges for the GO model39

Name Epsilon (Kcal/mole) Sigma (angstroms) Charge (e)CA, sp2 carbon 0.0700 3.5500 0.0000

CF, carbon in hydroxyl 0.0700 3.5500 0.1500

CT, carbon in epoxide 0.0660 3.5000 0.1400

HO, hydrogen in

hydroxyl

0.4600 0.4000 0.4350

OH, oxygen in hydroxyl 0.1700 3.0000 -0.5850

OS, oxygen in epoxide 0.1400 2.9000 -0.2800

References1 H. Huang, Y. Mao, Y. Ying, Y. Liu, L. Sun and X. Peng, Chem Commun (Camb), 2013, 49, 5963-

5965.2 W. Ying, J. Cai, K. Zhou, D. Chen, Y. Ying, Y. Guo, X. Kong, Z. Xu and X. Peng, ACS Nano, 2018,

12, 5385-5393.3 S. Plimpton, J. Comput. Phys., 1995, 117, 1-19.4 W. L. Jorgensen, D. S. Maxwell and J. Tirado-Rives, J. Am. Chem. Soc., 1996, 118, 11225-11236.5 W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. Merz, D. M. Ferguson, D. C. Spellmeyer,

T. Fox, J. W. Caldwell and P. A. Kollman, J. Am. Chem. Soc., 1995, 117, 5179-5197.6 J. de Andrade, E. S. Böes and H. Stassen, J. Phys. Chem. B, 2002, 106, 13344-13351.7 M. P. Allen and D. J. Tildesley, Computer simulation of liquids, Oxford university press, 2017.8 S. Jiao and Z. Xu, ACS Appl. Mater. Interfaces, 2015, 7, 9052-9059.9 C. D. Wick, T.-M. Chang and L. X. Dang, J. Phys. Chem. B, 2010, 114, 14965-14971.10 A. Mondal and S. Balasubramanian, J. Phys. Chem. B, 2016, 120, 4457-4466.11 S. Park, F. Khalili-Araghi, E. Tajkhorshid and K. Schulten, J. Chem. Phys., 2003, 119, 3559-3566.12 M. L. Connolly, J. Mol. Graph., 1993, 11, 139-143.13 T. H. Bae, J. S. Lee, W. Qiu, W. J. Koros, C. W. Jones and S. Nair, Angew. Chem. Int. Ed., 2010,

49, 9863-9866.14 A. K. Zulhairun and A. F. Ismail, J. Membr. Sci., 2014, 468, 20-30.15 S. Himeno, T. Tomita, K. Suzuki, K. Nakayama, K. Yajima and S. Yoshida, Ind. Eng. Chem. Res.,

2007, 46, 6989-6997.16 C. Kong, H. Du, L. Chen and B. Chen, Energy Environ. Sci., 2017, 10, 1812-1819.17 Z. Qiao, S. Zhao, M. Sheng, J. Wang, S. Wang, Z. Wang, C. Zhong and M. D. Guiver, Nat. Mater.,

2019, 18, 163-168.18 S. Duan, T. Kouketsu, S. Kazama and K. Yamada, J. Membr. Sci., 2006, 283, 2-6.19 S. Li, Z. Wang, X. Yu, J. Wang and S. Wang, Adv. Mater., 2012, 24, 3196-3200.20 Z. Qiao, Z. Wang, C. Zhang, S. Yuan, Y. Zhu, J. Wang and S. Wang, AlChE J., 2013, 59, 215-228.21 M. Wang, Z. Wang, S. Li, C. Zhang, J. Wang and S. Wang, Energy Environ. Sci., 2013, 6, 539-551.22 X. Jiang, S. Li and L. Shao, Energy Environ. Sci., 2017, 10, 1339-1344.23 Q. Song, S. K. Nataraj, M. V. Roussenova, J. C. Tan, D. J. Hughes, W. Li, P. Bourgoin, M. A. Alam,

A. K. Cheetham, S. A. Al-Muhtaseb and E. Sivaniah, Energy Environ. Sci., 2012, 5, 8359-8369.24 T. Rodenas, I. Luz, G. Prieto, B. Seoane, H. Miro, A. Corma, F. Kapteijn, F. X. Llabres i Xamena

and J. Gascon, Nat. Mater., 2015, 14, 48-55.25 J. E. Bachman and J. R. Long, Energy Environ. Sci., 2016, 9, 2031-2036.26 O. G. Nik, X. Y. Chen and S. Kaliaguine, J. Membr. Sci., 2012, 413, 48-61.27 W. Fam, J. Mansouri, H. Li, J. Hou and V. Chen, ACS Appl. Mater. Interfaces, 2018, 10, 7389-7400.28 J. Shen, G. Liu, K. Huang, W. Jin, K. R. Lee and N. Xu, Angew. Chem. Int. Ed., 2015, 54, 578-582.29 W. Fam, J. Mansouri, H. Li and V. Chen, J. Membr. Sci., 2017, 537, 54-68.30 M. Karunakaran, L. F. Villalobos, M. Kumar, R. Shevate, F. H. Akhtar and K. V. Peinemann, J.

Mater. Chem. A, 2017, 5, 649-656.31 M. Karunakaran, R. Shevate, M. Kumar and K. V. Peinemann, Chem. Commun., 2015, 51, 14187-

14190.32 H. W. Kim, H. W. Yoon, S. M. Yoon, B. M. Yoo, B. K. Ahn, Y. H. Cho, H. J. Shin, H. Yang, U.

Paik and S. Kwon, Science, 2013, 342, 91-95.

33 W. Zhao, G. He, L. Zhang, J. Ju, H. Dou, F. Nie, C. Li and H. Liu, J. Membr. Sci., 2010, 350, 279-285.34 D. Ji, Y. S. Kang and S. W. Kang, Sci. Rep., 2015, 5, 16362.35 Y. S. Park, C. Ha and S. W. Kang, RSC Adv., 2017, 7, 33568-33571.36 E. Ghasemi Estahbanati, M. Omidkhah and A. Ebadi Amooghin, ACS Appl. Mater. Interfaces, 2017,

9, 10094-10105.37 Y. Choi, G. H. Hong and S. W. Kang, J. Nanosci. Nanotechno., 2016, 16, 2832-2835.38 G. H. Hong, D. Ji and S. W. Kang, RSC Adv., 2014, 4, 16917.39 S. Safaei and R. Tavakoli, Desalination, 2017, 422, 83-90.