Ion-Gated Gas Separation through Porous Graphene - Nano Letters

Feb 23, 2017 - Department of Chemistry, University of California, Riverside, California ... Department of Chemistry, The University of Tennessee, Knox...
0 downloads 0 Views 983KB Size
Subscriber access provided by University of Newcastle, Australia

Communication

Ion-Gated Gas Separation through Porous Graphene Ziqi Tian, Shannon M. Mahurin, Sheng Dai, and De-en Jiang Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.6b05121 • Publication Date (Web): 23 Feb 2017 Downloaded from http://pubs.acs.org on February 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.

Nano Letters 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 14

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

Nano Letters

Ion-Gated Gas Separation through Porous Graphene Ziqi Tian,† Shannon M. Mahurin,‡ Sheng Dai,*,‡,§ and De-en Jiang *,† †



Department of Chemistry, University of California, Riverside, California 92521, US

Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, US §

Department of Chemistry, The University of Tennessee, Knoxville, Tennessee 37996, US *Corresponding authors: [email protected]; [email protected]

Abstract Porous graphene holds great promise as a one-atom-thin, high-permeance membrane for gas separation, but to precisely control the pore size down to three to five angstroms proves challenging. Here we propose an ion-gated graphene membrane comprising a monolayer of ionic liquid-coated porous graphene to dynamically modulate the pore size to achieve selective gas separation. This approach enables the otherwise non-selective large pores on the order of 1 nm in size to be selective for gases whose diameters range from three to four angstroms. We show from molecular dynamics simulations that CO2, N2 and CH4 all can permeate through a nanopore in graphene without any selectivity. But when a monolayer of [emim][BF4] ionic liquid (IL) is deposited on the porous graphene, CO2 has much higher permeance than the other two gases. We find that the anion dynamically modulates the pore size by hovering above the pore and provides affinity for CO2 while the larger cation (which cannot go through the pore) holds the anion in place via electrostatic attraction. This composite membrane is especially promising for CO2/CH4 separation, with a CO2/CH4 selectivity of about 42 and CO2 permeance of ~105 GPU (gas permeation unit). We further demonstrate that selectivity and permeance can be tuned by the anion size, pore size, and IL thickness. The present work points toward a promising direction of using the atom-thin ionic-liquid/porous-graphene hybrid membrane for high-permeance, selective gas separation that allows a greater flexibility in substrate pore size control. Keywords: Porous graphene, gas separation, ionic liquid, membrane, molecular dynamics

ACS Paragon Plus Environment

1

Nano Letters

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 14

Gas separation is of great importance in chemical and fuels production as well as in environmental remediation.1,

2

Removing acid gases, specifically carbon dioxide, is widely

practiced in natural gas processing and has become increasingly relevant for pre- and postcombustion carbon capture.3-5 Compared with cryogenic distillation and sorption, membrane technology does not require a phase change in the separation process, thereby reducing the costs of energy and environmental impact. Since the permeance is inversely proportional to the membrane thickness, single-layer graphene with well-defined nanoscale pores has been proposed as a one-atom-thin membrane for selective gas separation6-9 and water desalination.10-14 Simulations showed that the nanopores can separate gas molecules, for instance CO2, N2 and CH4, with high permeance and selectivity.6, 15-17 To achieve porous graphene for gas separation, ultraviolet-induced oxidative (UV/ozone) etching and ion/electron beam were used to create sub-nanometer pores in pristine graphene.18-20 And molecular sieving by the porous graphene membrane was demonstrated by several experiments.7, 21-24

However, it is still extremely difficult to fabricate defect-free porous membranes in large

scale that have precisely controlled pores matching the kinetic diameters of common gas molecules, ranging from 2.5 to 4.0 Å.25 Most of the synthesized porous-graphene membranes have pore sizes over 10 Å, while 2D porous polymers and covalent organic frameworks (COFs) prepared through the bottom-up method usually have pores of larger than 15 Å in size.26-29 To overcome the difficulty of pore-size control in one-atom-thin membranes, it would be highly desirable to be able to dynamically tune the pore size to smaller diameter for target gas separation, since for solid materials such as graphene the pore size is fixed once pores are created. It would be ideal to use the dynamics of a liquid to tune the pore size via the liquid-pore interaction. The liquid control layer needs to be thin and non-volatile because if it is too thick, the permeance will be low, dictated by the solubility and diffusivity of the gas in the liquid; if it is volatile, the liquid will be lost to evaporation. Due to their extremely low vapor pressure and great chemical tunability, room temperature ionic liquids (RTILs) have been used for gas separation both as a sorbent and in a supported membrane system.30-33 They are an ideal choice for the control layer. We design a composite membrane composed of a monolayer of ionic liquid atop a porous graphene with large pores (~1 nm) to demonstrate the idea of dynamically tuning the pore size for selective gas separation through classical molecular dynamics (CMD) simulations (see the

ACS Paragon Plus Environment

2

Page 3 of 14

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

Nano Letters

Methods section at the end and the Supporting Information, SI, for details). We choose porous graphene as a model ultrathin membrane due to its simple structure, while the porous substrate can also be extended to other 2D materials with nanopores, such as MoS2, 2D COFs and 2D polymers. The atomic thickness of the membrane enables high permeance, while the ionic liquid layer both attracts gas to the membrane surface and further modulates the size of the relatively large nanopore, resulting in high selectivity. In addition, the graphene layer can tailor the structure of the ultrathin ionic liquid layer for the benefit of gas adsorption and permeation. This synergistic interaction between graphene and ionic liquid is key to selective gas transport. A bi-chamber system is used to simulate CO2, N2 and CH4 permeation through the composite membrane from the feed side to the permeate side, as shown in Figure 1a. The porous graphene contains hydrogen-terminated hexagonal pores with a diameter of 6.0 Å (Figure 1b). This pore size can be more easily created experimentally than the smaller pores.6, 16 We then deposit a monolayer of [emim][BF4] ionic liquid on the feed side of the porous graphene; the average thickness of the ionic liquid (IL) layer is less than 5 Å. Wettability of ionic liquid on graphene has been investigated previously;34, 35 recent coarse-grained MD simulation showed that [emim][BF4] could form a single adsorbed layer film on a graphene sheet.35 Before investigating the performance of the composite membrane for gas separation, we first examine gas permeation through the porous graphene layer without the ionic liquid layer.

Figure 1. (a) The bi-chamber system for simulation of gas permeance through the composite membrane; top and bottom sides of the box are impermeable. (b) Structure of a pore inside the graphene layer with an electronic density contour. (c) Top view of the composite membrane from the feed side: a monolayer of

ACS Paragon Plus Environment

3

Nano Letters

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 4 of 14

the [emim][BF4] ionic liquid on the porous graphene. Color code: C, gray; H, white; N, blue; B, red; F, green. Cations are highlighted with blue edge, anions with red.

The hexagonal pore in the graphene layer (Figure 1b) has a diagonal C-C distance of 10 Å and H-H distance of 8.18 Å. Taking into account the van der Waals radius of hydrogen (1.2 Å), the estimated diameter of the pore is about 6 Å, which is much greater than the kinetic diameters of CO2 (3.30 Å), N2 (3.64 Å), and CH4 (3.80 Å).25 Thus without the IL layer, all three gases should pass through the porous graphene without any hindrance. Our all-atom classical CMD

simulations show that this is indeed the case (Figure 2). One can see that CH4, CO2 and N2 cross the porous graphene quickly and dynamic equilibriums are achieved after approximately 1 ns. The initial slopes correspond to a permeance on the order of 106 GPU with essentially no selectivity. In fact, the initial flux of CH4 is slightly higher than those of CO2 and N2, due to stronger adsorption of CH4 on the graphene surface.36

Figure 2. Pure gas permeation through the 6.0-Å porous graphene without the ionic liquid layer with an initial feed pressure of 10 atm at 298 K.

Next we examined how the monolayer IL can modulate gas transport through the pore graphene layer. After equilibration of the composite membrane system (Figure 1c), we found that both cations and anions stay on the feed side of the graphene membrane and neither can go through the pores. This is because the [emim]+ cations are too large to pass through the nanopores, while the smaller BF4- anions are held back by the electrostatic interaction from the cations. Radial distribution functions of ions around the pore centers (Figure 3) show that anions

ACS Paragon Plus Environment

4

Page 5 of 14

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

Nano Letters

are tightly interacting with the pores, due to the favorable interaction with the positively charged hydrogen atoms on the pore rim (Figure 1b). This can be clearly seen in a snapshot of the composite membrane in Figure 1c. We found that the centers of the anions are about 1.4 Å above the pore center.

Figure 3. Radial distribution function of the cations and anions around the pore center for a monolayer of the [emim][BF4] ionic liquid on the 6.0-Å porous graphene (Figure 1c).

Figure 4a shows the performance of the composite membrane for pure gas permeation. Indeed, one can see that the membrane is now highly selective for CO2 permeation, while CH4 permeation is reduced the most. Based on linear regression, the fluxes are 0.59, 0.12 and 0.014 molecules/ns for CO2, N2, and CH4, respectively, corresponding to permeance values of 1.39×105, 0.27×105 and 0.033×105 GPU (gas permeation unit; 1 GPU = 3.35×10-10 mol·m-2·s1

·Pa-1). These data lead to a very impressive pure-gas selectivity of 42 for CO2/CH4. For

comparison, the porous graphene alone with a 3.4-Å pore was shown to have a simulated CO2 permeance of 2.9×105 GPU and an estimated CO2/CH4 selectivity of 6.6×107.37 In another theoretical report, a 2D porous polymer membrane composed of stilbene units with a 3.5-Å pore had a simulated CO2 permeance of 3×105 GPU, with an estimated CO2/CH4 selectivity over 500.38 So a monolayer of ionic liquid on the porous graphene does not significantly impact its CO2 permeance. Of course, the 3.4-3.5 Å pore can offer much higher CO2/CH4 selectivity, but experimentally it is extremely challenging to precisely control the pore to this size in a 2D material in a scalable fashion.7 Effort to use scalable synthesis via the graphene oxide route

ACS Paragon Plus Environment

5

Nano Letters

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 6 of 14

generated membrane of a few nanometers in thickness and the measured membrane performance gave a CO2 of 102 GPU and CO2/CH4 selectivity of 10.22 On the other hand, state-of-the-art highpermeability polymeric membranes have CO2/CH4 permselectivity less than 40,39 while the conventional supported IL membranes currently have CO2/CH4 permselectivity less than 30.32

Figure 4. (a) Pure gas permeation through and (b) gas adsorption on the composite membrane of a monolayer of the [emim][BF4] ionic liquid on the porous graphene (Figure 1), with an initial feed pressure of 10 atm at 298 K.

To understand the origin of the permselectivity, we monitored the gas adsorption within the space of 10 Å above the graphene surface. As shown in Figure 4b, gas adsorption reaches equilibrium quite quickly (less than 1 ns). At equilibrium, the numbers of adsorbed molecules are 26.6, 5.84 and 9.71 for CO2, N2 and CH4, respectively, yielding an adsorption selectivity of 4.6 and 2.7 for CO2/N2 and CO2/CH4, respectively (Table 1). Thus, CO2/N2 permselectivity is mainly due to adsorption selectivity, meaning a more favorable interaction of the IL layer with CO2 than with N2. In contrast, CO2/CH4 permselectivity is mainly due to diffusivity selectivity (Table 1), which is closely related to how the ionic liquid layer modulates the pore size. Here we note that in Table 1 we assumed that permselectivity is a product of adsorption selectivity and diffusion selectivity, based on our observation of the formation of a gas adsorption layer on the 2D membrane (Figure 4b and more discussion below). For a more accurate treatment, one can introduce the Langmuir adsorption model to interpret the simulation data, as done previously.38

ACS Paragon Plus Environment

6

Page 7 of 14

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

Nano Letters

Table 1. Contributions to CO2/N2 and CO2/CH4 permselectivity from adsorption and diffusion. Gas pair

CO2/N2

CO2/CH4

Permselectivity

5.2

42

Adsorption selectivity

4.6

2.7

Diffusion selectivity

1.1

16

To further examine how the anion modulates the pore size to achieve selectivity and whether the modulation is affected by the gas molecule, we analyzed the distances between the pore center to the closest anion with and without CO2 (Figure S1 in SI). The shapes of two curves are quite similar, indicating that the dynamic modulation of the pore size is mainly due to the fluctuations in the ions. This is consistent with our finding that all three gases (CO2, N2, and CH4) can permeate through the composite membrane (Figure 4a) despite their different degrees of interaction with the ions (Figure 4b). On the other hand, Figure S1 does show that the presence of CO2 shifts the distribution curve to the right, thereby slightly increasing the gap between the ion and the pore. This observation is consistent with CO2’s higher permeance. We next examine the spatial distribution of ions and gases along the z-direction (membrane surface normal) on the feed side. Figure 5 shows that some anions are very close to the membrane (at z = 1.5 Å), while the main anion and cation layers are at z = 4.0 and 3.6 Å, respectively. One can clearly see the adsorbed gas layer (z = 6~9 Å) on top of the ion layer (z = 3~5 Å). In addition to these main peaks, one can also see the close interaction of CO2 inside the IL layer (z = 3.2 Å). To better show the spatial position of gas molecules relative to the pore center and the graphene layer, we plotted the probability densities of the three gas in a 2D (r,z) contour plot (Figure 6), where r is the distance of the gas to the pore center and z, again, is the distance of the gas to the graphene layer. One can clearly see some probability of CO2 close to the pore center (z = 1.5~3 Å) and also inside the pore (z = 0~1 Å), in addition to the main adsorbed layer at z = 6~9 Å; in contrast, N2 and CH4 have much less probability on the liquid layer and no significant probability close to the pore.

ACS Paragon Plus Environment

7

Nano Letters

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 14

Figure 5. Probability densities of the ions (left axis) and the gases (right axis) away from the porous graphene layer along the surface normal (z-direction) on the feed side, for a monolayer of [emim][BF4] on the 6.0-Å porous graphene (Figure 1c). Graphene layer is at z=0.

Figure 6. 2D contour plot of probability densities of the gas molecule on the composite membrane of a monolayer of [emim][BF4] on the 6.0-Å porous graphene (Figure 1c); r is the distance of the gas to the pore center and z is the distance of the gas to the graphene layer.

To further reveal the atomistic details of CO2 permeation, we show three representative snapshots of a CO2 passing-through event in Figure 7. First, the CO2 molecule is attracted by the anion to the pore center. Then CO2 moves around the anion and squeezes into the gap between the anion and the pore rim. Next, CO2 nudges the anion off the pore center and passes through

ACS Paragon Plus Environment

8

Page 9 of 14

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

Nano Letters

the pore. After CO2 crosses the membrane, the anion relaxes back to the center, blocking the pore again. Nitrogen has a similar size to CO2, so it can also cross the membrane by the same process, but its low surface concentration leads to a lower flux. In contrast, it is much more difficult for CH4 (which has a much larger diameter) to squeeze pass the anion blocking the pore.

Figure 7. Snapshots of a passing-through event of CO2 through the composite membrane: approaching (left), entering (middle), and exiting (right) the pore. Color code: C, cyan; H, white; N, blue; O, red; B, pink; F, red purple.

Since the anion-pore interaction is a key factor in modulating the pore size, we explored the ion size effect by changing [BF4]- to a larger anion, [PF6]-. We found that the permeances of both CO2 and N2 decrease, with N2 showing a greater decrease, resulting in an increase in the CO2/N2 permselectivity for [emim][PF6] compared to [emim][BF4]. Because the adsorption selectivity slightly decreases, the increase in the CO2/N2 permselectivity from [BF4]- to [PF6]- is therefore mainly due to the diffusivity selectivity. In other words, the effective pore size can be tuned by varying the anion. For CH4, no passing-through event was observed for the [emim][PF6]/porous-graphene membrane in the 50-ns MD simulation. Next we investigated how the graphene pore size and the thickness of the IL layer affect the gas permeance. We found that if a larger pore (diameter 9.6 Å; Figure S2a) is used with the [emim][BF4] monolayer, the CO2/CH4 selectivity decreases to 1 (Figure S2c) because now even the ions can permeate through the pore. We went a step further removing the porous-graphene layer and found that the hypothetical IL monolayer does not have CO2/CH4 selectivity, either (Figure S3). On the other hand, when we used a smaller pore (diameter 4.2 Å; Figure S2b) with the [emim][BF4] monolayer, the flux of CO2 decreased by 60 times compared to that for the 6.0Å pore, even though the 4.2-Å pore itself without the IL present could allow both CO2 and CH4 to pass with high permeance and no selectivity. We further tested the 6.0-Å porous graphene (Figure 1b) with two layers of the [emim][BF4] IL. The results (Figure S4) show that the flux of

ACS Paragon Plus Environment

9

Nano Letters

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 10 of 14

CO2 decreased by 75% from the monolayer IL to the bilayer, even though the bilayer composite membrane was still selective for CO2/CH4 separation. After considering the different pore sizes, ion sizes, and IL thickness, we can now conclude that the monolayer coating is ideal in optimizing both the selectivity and permeance; the porous graphene layer is not merely a support, but it is the matching between the pore size and the ion size that yields both high selectivity and high permeance. In the pores that we examined here (Figure 1c and Figure S2), the pore rim has positive charges due to hydrogen termination, so it attracts the anion. The pore rim can also be functionalized with electronegative dopants such as nitrogen, so cations can be also used to modulate the pore size. Moreover, the metallic feature of the graphene layer can be used to electrochemically modulate charges in the graphene layer and the graphene-IL interaction, as explored in the study of electric double-layer capacitators. In this case, the metallic nature of the graphene layer becomes important and one needs to take into account the charge fluctuation in the graphene layer in response to the dynamics of the IL layer by using either first principles methods or the polarizable force fields.40-42 To realize the IL/porous-graphene membrane for gas separation, the following experimental steps are suggested. First, large area graphene with no defects can be grown using chemical vapor deposition on a copper foil catalyst either in a batch process or roll-to-roll.43 Next, a polymer transfer layer, e.g., poly(methyl methacrylate) or PMMA, is coated onto the graphene/copper system and the copper is then removed.14 The PMMA/graphene is then transferred to the desired substrate and the PMMA is removed. Once the graphene is transferred to a porous substrate, small pores can be introduced through oxygen plasma etching or ion bombardment of chemical etching.18-20 A thin layer of ionic liquid can be drop cast or spin coated from a suitable solvent such as ethanol. The obtained membrane can then be used for testing gas permeation. Various challenges can arise in the procedures above, for example, avoiding tears during the graphene transfer and control of the ionic-liquid thickness. In summary, we demonstrate from classical molecular dynamics simulations that a monolayer of ionic liquid can be used to control the pore sizes of a porous graphene that otherwise would be too large for selective gas separation. The IL-on-graphene membrane combines many unique features such as cation-anion interaction, ion-graphene interaction, anion-

ACS Paragon Plus Environment

10

Page 11 of 14

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

Nano Letters

pore interaction, gas-anion interaction, dynamic pore modulation, atom-thin membrane thickness, and high permeance. In the specific example of the [emim][BF4] ionic liquid on a hydrogenterminated porous graphene, we found that the pore is gated by the strongly interacting anion, which renders the pore highly selective for CO2/CH4 separation and achieves a selectivity over 40 and CO2 permeance on the order of 105 GPU. Without the IL layer, the porous graphene is not selective for CO2/CH4 separation. We further showed that the selectivity can be tuned by varying the anion using the CO2/N2 separation as an example. This work suggests a new avenue of research to achieve dynamic control of pore size in one-atom-thin porous membranes for highpermeance gas separation. Methods. All the CMD simulations were carried out with the LAMMPS package.44 The nonpolarizable OPLS-AA type force field45 was employed for the ionic liquids; all the parameters were provided in SI. The graphene layer was fixed in the center of the simulation box. Atomic partial charges on porous graphene was determined based on the electrostatic potential.46 For CO2 and N2, three-site models were adopted as in our previous simulations;47 all-atom model was used for methane. Lennard-Jones potential terms were evaluated via the Lorentz-Berthelot mixing rule with a cutoff of 12 Å. All parameters for the gases and porous graphene were also provided in SI. Periodic boundary condition was applied in x and y directions and the long-range electrostatic interaction was evaluated via the 3D Particle-Particle-Particle-Mesh (PPPM) algorithm with a slab correction for the main simulation runs. All the data reported in this work were averaged over a set of 20 simulations from various initial velocity distributions. For each simulation, the ionic liquid layer was initially heated up to 1000 K for 1 ns and quenched to 300 K in 1 ns, to coat the ionic liquid on the porous graphene evenly. In most cases, ionic liquid formed a uniform layer, covering all the nanopores. After ionic liquid coating, 25 ns simulation was carried out in NVT ensemble. The temperature of the fluid was kept at 300 K with the NoseHoover algorithm.48, 49 More computational details are provided in SI.

Notes The authors declare no competing financial interest.

ACS Paragon Plus Environment

11

Nano Letters

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 12 of 14

Acknowledgement. This work was supported by the Division of Chemical Sciences, Geosciences and Biosciences, Office of Basic Energy Sciences, U.S. Department of Energy. This research used resources of the National Energy Research Scientific Computing Center, a DOE Office of Science User Facility supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. Supporting Information Available: • Computational details. • Simulation results of different pore sizes, ionic liquid thickness, and analysis of the pore-ion interaction. • All force-field parameters.

Reference (1) Bernardo, P.; Drioli, E.; Golemme, G. Ind. Eng. Chem. Res. 2009, 48, 4638-4663. (2) Baker, R. W.; Low, B. T. Macromolecules 2014, 47, 6999-7013. (3) Merkel, T. C.; Lin, H.; Wei, X.; Baker, R. J. Membr. Sci. 2010, 359, 126-139. (4) Tian, Z. Q; Dai, S.; Jiang, D. E. WIREs Comput. Mol. Sci., 2016, 6, 173-197. (5) Boot-Handford, M. E.; Abanades, J. C.; Anthony, E. J.; Blunt, M. J.; Brandani, S.; Mac Dowell, N.; Fernandez, J. R.; Ferrari, M.-C.; Gross, R.; Hallett, J. P., et al. Energy Environ. Sci., 2014, 7, 130-189. (6) Jiang, D. E.; Cooper, V. R.; Dai, S. Nano Lett. 2009, 9, 4019-4024. (7) Koenig, S. P.; Wang, L. D.; Pellegrino, J.; Bunch, J. S. Nat. Nanotechnol. 2012, 7, 728-732. (8) Wang, L. D.; Drahushuk, L. W.; Cantley, L.; Koenig, S. P.; Liu, X. H.; Pellegrino, J.; Strano, M. S.; Bunch, J. S. Nat. Nanotechnol. 2015, 10, 785-790. (9) Huang, L.; Zhang, M.; Li, C.; Shi, G. Q. J. Phys. Chem. Lett. 2015, 6, 2806-2815. (10) Sint, K.; Wang, B.; Kral, P. J. Am. Chem. Soc. 2008, 130, 16448-16449. (11) Cohen-Tanugi, D.; Grossman, J. C. Nano Lett. 2012, 12, 3602-3608. (12) Wang, E. N.; Karnik, R. Nat. Nanotechnol. 2012, 7, 552-554. (13) O'Hern, S. C.; Boutilier, M. S. H.; Idrobo, J. C.; Song, Y.; Kong, J.; Laoui, T.; Atieh, M.; Karnik, R. Nano Lett. 2014, 14, 1234-1241. (14) Surwade, S. P.; Smirnov, S. N.; Vlassiouk, I. V.; Unocic, R. R.; Veith, G. M.; Dai, S.; Mahurin, S. M. Nat. Nanotechnol. 2015, 10, 459-464. (15) Blankenburg, S.; Bieri, M.; Fasel, R.; Mullen, K.; Pignedoli, C. A.; Passerone, D. Small 2010, 6, 2266-2271. (16) Liu, H.; Dai, S.; Jiang, D. E. Nanoscale 2013, 5, 9984-9987. (17) Liu, H. J.; Chen, Z. F.; Dai, S.; Jiang, D. E. J. Solid State Chem. 2015, 224, 2-6. (18) Liu, L.; Ryu, S. M.; Tomasik, M. R.; Stolyarova, E.; Jung, N.; Hybertsen, M. S.; Steigerwald, M. L.; Brus, L. E.; Flynn, G. W. Nano Lett. 2008, 8, 1965-1970. (19) Russo, C. J.; Golovchenko, J. A. Proc. Natl. Acad. Sci. 2012, 109, 5953-5957.

ACS Paragon Plus Environment

12

Page 13 of 14

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

Nano Letters

(20) Celebi, K.; Buchheim, J.; Wyss, R. M.; Droudian, A.; Gasser, P.; Shorubalko, I.; Kye, J. I.; Lee, C.; Park, H. G. Science 2014, 344, 289-292. (21) O'Hern, S. C.; Stewart, C. A.; Boutilier, M. S. H.; Idrobo, J. C.; Bhaviripudi, S.; Das, S. K.; Kong, J.; Laoui, T.; Atieh, M.; Karnik, R. ACS Nano 2012, 6, 10130-10138. (22) Kim, H. W.; Yoon, H. W.; Yoon, S. M.; Yoo, B. M.; Ahn, B. K.; Cho, Y. H.; Shin, H. J.; Yang, H.; Paik, U.; Kwon, S.; Choi, J. Y.; Park, H. B. Science 2013, 342, 91-95. (23) Joshi, R. K.; Carbone, P.; Wang, F. C.; Kravets, V. G.; Su, Y.; Grigorieva, I. V.; Wu, H. A.; Geim, A. K.; Nair, R. R. Science 2014, 343, 752-754. (24) O'Hern, S. C.; Jang, D.; Bose, S.; Idrobo, J. C.; Song, Y.; Laoui, T.; Kong, J.; Karnik, R. Nano Lett. 2015, 15, 3254-3260. (25) Mehio, N.; Dai, S.; Jiang, D. E. J. Phys. Chem. A 2014, 118, 1150-1154. (26) Spitler, E. L.; Dichtel, W. R. Nat. Chem. 2010, 2, 672-677. (27) Colson, J. W.; Dichtel, W. R. Nat. Chem. 2013, 5, 453-465. (28) Huang, N.; Chen, X.; Krishna, R.; Jiang, D. L. Angew. Chem., Int. Ed. 2015, 54, 2986-2990. (29) Bisbey, R. P.; DeBlase, C. R.; Smith, B. J.; Dichtel, W. R. J. Am. Chem. Soc. 2016, 138, 1143311436. (30) Krull, F.; Fritzmann, C.; Melin, T. J. Membr. Sci. 2008, 325, 509-519. (31) Uchytil, P.; Schauer, J.; Petrychkovych, R.; Setnickova, K.; Suen, S. Y. J. Membr. Sci. 2011, 383, 262-271. (32) Tome, L. C.; Marrucho, I. M. Chem. Soc. Rev. 2016, 45, 2785-2824. (33) Dai, Z. D.; Noble, R. D.; Gin, D. L.; Zhang, X. P.; Deng, L. Y. J. Membr. Sci. 2016, 497, 1-20. (34) Herrera, C.; García, G.; Atilhan, M.; Aparicio, S. J. Phys. Chem. C 2015, 119, 24529-24537. (35) Burt, R.; Birkett, G.; Salanne, M.; Zhao, X. S. J. Phys. Chem. C 2016, 120, 15244-15250. (36) Sun, C.; Boutilier, M. S.; Au, H.; Poesio, P.; Bai, B.; Karnik, R.; Hadjiconstantinou, N. G. Langmuir 2014, 30, 675-682. (37) Liu, H. J.; Chen, Z. F.; Dai, S.; Jiang, D. E. J. Solid State Chem. 2015, 224, 2-6. (38) Schrier, J. ACS Appl. Mater. Interfaces 2012, 4, 3745-3752. (39) Wang, S. F.; Li, X. Q.; Wu, H.; Tian, Z. Z.; Xin, Q. P.; He, G. W.; Peng, D. D.; Chen, S. L.; Yin, Y.; Jiang, Z. Y.; Guiver, M. D. Energy Environ. Sci. 2016, 9, 1863-1890. (40) Vatamanu, J.; Borodin, O.; Smith, G. D. J. Am. Chem. Soc. 2010, 132, 14825-14833. (41) Limmer, D. T.; Merlet, C.; Salanne, M.; Chandler, D.; Madden, P. A.; van Roij, R.; Rotenberg, B. Phys. Rev. Lett. 2013, 111, 106102. (42) Merlet, C.; Pean, C.; Rotenberg, B.; Madden, P. A.; Simon, P.; Salanne, M. J. Phys. Chem. Lett. 2013, 4, 264-268. (43) Li, X. S.; Colombo, L.; Ruoff, R. S. Adv. Mater. 2016, 28, 6247-6252. (44) Plimpton, S. J. Comput. Phys. 1995, 117, 1-19. (45) Jorgensen, W. L.; Maxwell, D. S.; TiradoRives, J. J. Am. Chem. Soc. 1996, 118, 11225-11236. (46) Campana, C.; Mussard, B.; Woo, T. K. J. Chem. Theory Comput. 2009, 5, 2866-2878. (47) Liu, H. J.; Dai, S.; Jiang, D. E. J. Phys. Chem. B 2014, 118, 2719-2725. (48) Nose, S. J. Chem. Phys. 1984, 81, 511-519. (49) Hoover, W. G. Phys. Rev. A 1985, 31, 1695-1697.

ACS Paragon Plus Environment

13

Nano Letters

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 14 of 14

TOC Graphic

ACS Paragon Plus Environment

14