Water Desalination through Zeolitic Imidazolate Framework

Nov 20, 2015 - The five ZIFs possess identical rho-topology but differ in functional groups. The rejection of salt (NaCl) is found to be around 97% in...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Water Desalination through Zeolitic Imidazolate Framework Membranes: Significant Role of Functional Groups Krishna M. Gupta, Kang Zhang, and Jianwen Jiang* Department of Chemical and Biomolecular Engineering, National University of Singapore, 117576, Singapore S Supporting Information *

ABSTRACT: A molecular simulation study is reported for water desalination through five zeolitic imidazolate framework (ZIF) membranes, namely ZIF-25, -71, -93, -96, and -97. The five ZIFs possess identical rho-topology but differ in functional groups. The rejection of salt (NaCl) is found to be around 97% in ZIF-25, and 100% in the other four ZIFs. The permeance ranges from 27 to 710 kg/(m2·h·bar), about one∼two orders of magnitude higher compared with commercial reverse osmosis membranes. Due to a larger aperture size da, ZIF-25, -71, and -96 exhibit a much higher water flux than ZIF-93 and -97; however, the flux in ZIF-25, -71, and -96 is governed by the polarity of functional group rather than da. With the hydrophobic CH3 group, ZIF-25 has the highest flux despite the smallest da among ZIF-25, -71, and -96. The lifetime of hydrogen bonding in ZIF-25 is shorter than that in ZIF-71 and -96. Furthermore, water molecules undergo a fast flushing motion in ZIF-25, but frequent jumping in ZIF-96 and particularly in ZIF-97. An Arrhenius-type relationship is found between water flux in ZIF-25 and temperature, and the activation energy is predicted to be 6.5 kJ/mol. This simulation study provides a microscopic insight into water desalination in a series of ZIFs, reveals the key factors (aperture size and polarity of functional group) governing water flux, and suggests that ZIF-25 might be an interesting reverse osmosis membrane for high-performance water desalination.

1. INTRODUCTION Due to the rapidly growing population, energy demand, and industrialization, freshwater scarcity has become a major global concern.1,2 As over 95% of water on the Earth is seawater, there is considerable interest to desalinate seawater to supply freshwater.3,4 A handful of desalination techniques have been developed including reverse osmosis (RO) and multistage flash distillation (MSFD).5 Compared with thermal-based MSFD, RO is more energy efficient. Presently, RO produces daily 42 million kg of potable water, which accounts for 60% of world desalination capacity.6 The membrane in a RO process plays a pivotal role in desalination performance.7 Currently, polymeric membranes are commercially used for RO. Nevertheless, they suffer from drawbacks such as oxidation, fouling, and abrasion, thus leading to low membrane stability and insufficient water flux. The capital cost of current RO technology is high, and it needs to be improved prior to propagation. Toward this end, a larger number of studies have been conducted aiming to develop novel RO membranes including zeolitic and carbonaceous materials. With uniform pore structure and high thermal stability, zeolite membranes (e.g., MFI) exhibit high water flux and good salt rejection; however, they are fragile and not easily reproducible due to crack formation.8,9 Carbon nanotubes (CNTs) and graphene demonstrate good potential in water desalination, whereas their durability and scalability remain a practical bottleneck for large-scale industrial application.10−12 © XXXX American Chemical Society

In the past decade, metal−organic frameworks (MOFs) have emerged as a new class of porous materials.13 The judicious selection of inorganic and organic building blocks allows the pore size, volume, and functionality in MOFs to be tailored in a rational way. Consequently, MOFs have been considered versatile materials for storage, separation, catalysis, and many other potential applications.14 However, most current experimental and theoretical studies for MOFs have been focused on gas storage and separation, particularly the storage of lowcarbon-footprint energy carriers (e.g., H2 and CH4) and the separation of CO2-containing gas mixtures.15−21 As a proof-ofconcept, our group reported the first simulation study for water desalination through a MOF membrane.22 Specifically, the MOF membrane examined, zeolitic imidazolate framework (ZIF)-8, was predicted to be superior for RO. This simulation prediction was recently supported by experiment, in which 0.4% of ZIF-8 was added into a polyamide membrane and found to increase water permeance by 162%.23 As a subset of MOFs, ZIFs possess exceptional chemical and thermal stability; moreover, the pore size and affinity are readily tunable. Therefore, there is a large opportunity to explore ZIFs with high desalination performance. Received: September 25, 2015 Revised: November 8, 2015

A

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 1. Atomic structures of ZIF-25, -71, -93, -96, and -97. ZnN4 cluster: orange polyhedron, C: cyan, O: red, N: blue, Cl: green, and H: white.

The periodic boundary conditions were applied in x and y directions; thus, the ZIF membrane was mimicked to be infinitely large. The framework atoms of ZIFs were represented by Lennard−Jones (LJ) and electrostatic potentials

In this study, a series of ZIFs are investigated by simulations as potential RO membranes for seawater desalination. The ZIFs include ZIF-25, -71, -93, -96, and -97 with identical rhotopology, but differing in imidazolate linkers with various functional groups. Thus, the role of functional groups in water desalination through ZIF membranes can be exclusively revealed, which has not been reported in the literature. Following this introduction, the models of the five ZIFs and seawater, and the methods used, are briefly described in section 2. In section 3, salt rejection and water permeance/permeability through the ZIF membranes are presented and compared with other membranes, the structural and dynamic properties of water in the membranes are discussed, and the effects of pressure and temperature on water transport are also examined. Finally, the concluding remarks are summarized in section 4.

Unonbonded

⎡⎛ ⎞12 ⎛ ⎞6 ⎤ σij σij = ∑ 4εij⎢⎢⎜⎜ ⎟⎟ − ⎜⎜ ⎟⎟ ⎥⎥ + r ⎝ rij ⎠ ⎦ ⎣⎝ ij ⎠

∑

qiqj 4πε0rij

(1)

where εij and σij are the well depth and collision diameter, rij is the distance between atoms i and j, qi is the atomic charge of atom i, and ε0 = 8.8542 × 10−12 C2 N−1 m−2 is the permittivity of vacuum. The LJ parameters as listed in Table S1 were taken from the DREIDING force field,27 which is commonly used to describe ZIFs.28−30 The atomic charges of ZIFs were calculated using density functional theory (DFT) based on fragmental clusters (see Figure S1). The DFT calculations used the Becke exchange plus Lee−Yang−Parr functional (B3LYP) and were carried out using GAUSSIAN 09.31 The 6-31G(d) basis set was used for all the atoms of ZIFs except Zn atoms, for which the LANL2DZ basis set was used. By fitting the electrostatic potentials, the atomic charges were estimated as listed in Table S2. Water was modeled by the TIP3P,32 and Na+ and Cl− ions were described as charged LJ particles with potential parameters from the AMBER force field.33 The carbon atoms in graphene plates were mimicked by LJ potential with parameters as used for CNTs.34 Each system was initially subjected to energy minimization using the steepest descent method, and then velocities were assigned according to the Maxwell−Boltzmann distribution at 300 K. Finally, molecular dynamics (MD) simulation was conducted at 300 K with Pleft = 60.1 MPa and Pright = 0.1 MPa (ambient pressure). To estimate the activation energy of water transport, the system with ZIF-25 membrane was also run at 280, 290, 310, and 320 K. The temperatures were controlled by the velocity-rescaled Berendsen thermostat with a relaxation time of 0.1 ps. Moreover, Pleft in the system with ZIF-25 membrane varied from 30.1 to 60.1 and 90.1 MPa to investigate the effect of pressure. It is noteworthy that the pressures exerted were approximately 1 order of magnitude higher than practical values. This is common in nonequilibrium MD simulations, in order to reduce thermal noise and enhance signal/noise ratio within a nanosecond time scale currently affordable by most MD simulations. For instance, extremely high pressures (up to 600 MPa) were used to simulate water desalination through carbonaceous materials (CNTs, graphene, and graphyne).35−38 All the ZIF membranes were assumed to be rigid during simulations. A cutoff of 14 Å was used to calculate the LJ interactions, and the particle-mesh Ewald method was used to evaluate the electrostatic interactions with grid spacing of 1.2 Å and real-space cutoff of 14 Å. For each system, three independent simulations were performed each

2. MODELS AND METHODS Figure 1 shows the atomic structures of ZIF-25, -71, -93, -96, and -97. The metal clusters (ZnN4) in the five ZIFs are the same, but imidazolate linkers are different. The linkers are dimethyl imidazolate (dmeIm), dichloro imidazolate (dcIm), aldehydemethyl imidazolate (almeIm), cyanide amine imidazolate (cyamIm), and hydroxymethyl imidazolate (hymeIm) in ZIF-25, -71, -93, -96 and -97, respectively. The five ZIFs belong to the rho-topology, and the linkers are dually functionalized at positions 4 and 5. The pore sizes along the (001) direction were estimated from the Zeo++ program.24 Table 1 lists the Table 1. Structural Characteristics of ZIF-25, -71, -93, -96, and -97

diameters of cage (dc) and aperture (da) in each ZIF. The dc ranges marginally from 15.8 to 17.0 Å. The largest da is 5.5 Å in ZIF-96, and the smallest in ZIF-97 (3.5 Å) and -93 (3.7 Å). As we shall see below, the da plays a key role in governing water transport through the ZIF membranes. Water desalination through each ZIF membrane was simulated in a system schematically illustrated in Figure 2. There were two chambers containing NaCl solution and pure water, respectively. The concentration of NaCl was 0.5 M, close to the salt concentration in seawater. The two chambers were separated by a ZIF membrane with a thickness of one unit cell (28.6 Å). Additionally, two graphene plates were added into the chambers, and they could self-adjust their positions during simulation under hydraulic pressures Pleft and Pright, respectively. B

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 2. Simulation system for water desalination through a ZIF membrane. An aqueous NaCl solution with 0.5 M NaCl and pure water bath are on the left and right chambers of the membrane, respectively. Two graphene plates in the two chambers are exerted under hydraulic pressures Pleft and Pright, respectively. Zn: orange, N: blue, C: cyan, H: white; graphene: cyan; Na+: blue, Cl−: green; water molecules in the left and right chambers: yellow and magenta.

Figure 3. Number distributions of Na+ and Cl− ions at 12 ns. Each membrane is between the two dotted lines (z = 93 and 122 Å) and ΔP = 60 MPa.

systems at 12 ns (the final stage). Apparently, most Na+ and Cl− remain in the left chamber (NaCl solution). While a few ions cross the solution/membrane interface (z = 93 Å) because there are open cages along the membrane surface, they cannot enter the interior of ZIF-71, -93, and -97 membranes due to the restriction of small aperture. Nevertheless, ZIF-96 contains the largest aperture, as well as hydrophilic functional groups favoring interaction with ions; thus, both Na+ and Cl− ions are observed to enter ZIF-96 and reside therein. Moreover, the number of Na+ in ZIF-96 is higher than Cl− because bulkier Cl− (4.417 Å) encounters a larger steric repulsion compared with Na+ (3.238 Å). Interestingly, several Na+ ions pass through ZIF-25 and stay in the right chamber. This is attributed to the fast water transport in ZIF-25, as further discussed below, which facilitates ion transport. To examine the time-evolution of ion transport, Figure S2 plots the number distributions of Na+ and Cl− in the system with ZIF-25 at 3, 7, and 12 ns, respectively. As time lapses, the distributions of both Na+ and Cl− shift from the left to right and the peaks become higher. This is because RO occurs, water flows from the left chamber to right, and the volume of the left chamber is reduced. Consequently, ions are accumulated toward the membrane surface. Figure S3 shows the number of Na+ ions crossing the ZIF membranes into the right chamber. Obviously, Na+ ions can cross only ZIF-25 but not the other four ZIFs. The salt rejection was calculated by (cleft − cright)/cleft × 100%, where cleft and cright are the concentrations of Na+ ions in the left and right chambers, respectively. On this basis, the salt rejection by ZIF-25 is approximately 97% and 100% by ZIF-71, -93, -96, and -97. We should note that the

with 12 ns and found to give close results. The time step was 2 fs, and the trajectory was saved with a time interval of 2 ps. After 12 ns, the left and right compartments in each system were removed; then another 10 ns MD simulation was conducted, without adding hydraulic pressure, to examine the structural and dynamic properties of water in ZIF membranes. All the simulations used GROMACS package v.4.5.339 and analyzed by an in-house code.

3. RESULTS AND DISCUSSION First, the performances of water desalination through the five ZIF membranes are quantified on the basis of salt rejection and water permeance/permeability, and compared with other membranes. Then the structural and dynamic properties of water in the membranes are presented, including radial distribution functions and hydrogen bonds. Finally, the effects of pressure and temperature on water transport through ZIF-25 membrane are examined. 3.1. Salt Rejection and Water Permeance/Permeability through ZIFs. Upon the initiation of simulation, water in two chambers can exchange. Due to the existence of pressure gradient ΔP = Pleft − Pright, however, more water flows from the left to right. Consequently, the volume of the right chamber increases; meanwhile, the two graphene plates move from the left to right. This implies that RO occurs with net water transferred from NaCl solution to pure water bath. A movie is provided in the Supporting Information to visualize the RO process in the ZIF-96 membrane. Along with water transport, Na+ and Cl− ions also move toward the membrane, Figure 3 shows their number distributions along the z-axis in the five C

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 4. (a) Number of net transferred water molecules. (b) Water flux through ZIF membranes at ΔP = 60 MPa.

Figure 5. Performance: (a) permenace and (b) permeability vs salt rejection of ZIF and other membranes (MFI zeolite and commercial RO,7 functionalized and pristine CNTs,10 graphene,37 graphyne,38 and UiO-6640).

Figure 6. Radial distribution functions of water around the framework atoms of ZIF-25, -71, and -96. The notations of the framework atoms are in Figure S1.

and -97 (3.4 Å). However, the Jw through ZIF-25, -71, and -96 do not follow the increasing trend of da. Among these three, ZIF-25 possesses the smallest da (5.1 Å), but its Jw is the highest; the reverse is observed in ZIF-96 with the lowest Jw despite the largest da (5.5 Å). This implies that Jw is also governed by other factors, in addition to aperture size, as elucidated in section 3.2. Figure 5a illustrates the performances of ZIF-25, -71, -93, -96, -97, and other membranes in terms of salt rejection and water permeance (= Jw/ΔP). The five ZIF membranes, as well as ZIF-8 from our previous study,22 exhibit significantly higher salt rejection and water permeance than MFI zeolite. Compared with commercial seawater RO, brackish RO, and

ZIFs were assumed to be rigid in this study. If the framework flexibility is incorporated, the salt rejection might be affected. Figure 4a shows the number of net transferred water molecules Nw through ZIF-25, -71, -93, -96, and -97 at ΔP = 60 MPa. In each ZIF, the Nw increases almost linearly with time despite small fluctuations due to random thermal motion. From the slope Nw, water flux Jw was calculated by Jw = Nw/(At), where A is the membrane area and t is time duration. As shown in Figure 4b, the hierarchy of Jw is ZIF-25 > -71 > -96 > -93 > -97. Particularly, ZIF-25, -71, and -96 exhibit Jw substantially higher than that of ZIF-93 and -97. This is because a clear threshold exists in terms of the aperture size da. The da in ZIF25, -71, and -96 are >5 Å, larger than those in ZIF-93 (3.7 Å) D

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

forms approximately 2.7 hydrogen bonds in the three ZIFs, which is close to that in ZIF-8,22 but lower than the value (3.5) in bulk water.32 This suggests that the entry of water from bulk phase into the ZIF membrane is energetically unfavorable; nevertheless, the energy loss of weakened hydrogen bonding in the membrane is compensated by the interaction with framework. The dynamics of hydrogen bonding in the membranes is further explored using the autocorrelation function (ACF)44

high-flux RO membranes, the ZIFs are theoretically predicted to possess one∼two orders of magnitude higher permeance, though salt rejection is similar. Particularly, ZIF-25 membrane has excellent performance with permeance slightly lower than that of functionalized CNT,10 graphene,37 and graphyne.38 If membrane thickness S is taken into account, the permeability (Jw S/ΔP) in ZIF-25 is even higher than graphene and graphyne but close to CNT. The carbon-based CNT membranes, however, are currently difficult to be produced in a large scale with high quality. UiO-66 is also a MOF, but its performance is inferior to the ZIFs.40 3.2. Water Structure and Dynamics in ZIFs. From Figure 4, water flux Jw through ZIF-25, -71, and -96 does not follow the increasing trend of da. To unravel this unique behavior, water structure in these three ZIFs is characterized by radial distribution functions g(r) defined as gij(r ) =

c(t ) =

(3)

where h(t) = 1 if two water molecules are hydrogen bonded at time t and h(t) = 0 otherwise. The ensemble average ⟨···⟩ is on all the pairs of hydrogen bonded water molecules. c(t) is the probability that two water molecules remain hydrogen bonded at time t = 0 as well as time t or, in other words, indicates the time duration of hydrogen bonding. Physically, c(t) measures how fast a hydrogen bond is relaxed. When time approaches infinity, c(t) approaches zero. The lifetime of hydrogen bonding τHB is estimated by c(t = τHB) = e−1.45 As shown in Figure 7,

Nij(r , r + Δr )V 4πr 2ΔrNN i j

⟨h(0)h(t )⟩ ⟨h⟩

(2)

where r is the distance between atoms i (water molecules) and j (framework atoms), Nij(r,r + Δr) is the number of atom j around i within a shell from r to r + Δr, V is the volume of framework (i.e., membrane), and Ni and Nj are the numbers of atoms i and j, respectively. Figure 6 shows the g(r) of water around the framework atoms of ZIF-25, -71, and -96. While a marginal peak at 4 Å is observed in ZIF-25, there is a small peak at 3.7 Å in ZIF-71. Nevertheless, the g(r) around the N2 and N3 atoms in ZIF-96 have pronounced peak at 3 Å. This indicates that the framework affinity for water increases as ZIF25 < -71 < -96. Figure S4 shows the g(r) between water molecules in the three ZIFs. The peak of g(r) drops with increasing framework affinity. In ZIF-96 with high affinity, water is strongly bound onto the framework; as a consequence, water−water interaction is weak and the peak of g(r) is low. In ZIF-25, however, water molecules possess strong interaction among themselves and form clusters, showing a high peak in g(r). All these observations are attributed to the presence of different functional groups. ZIF-25 and -71 contain nonpolar (CH3) and weakly polar (Cl) groups; however, polar NH2 and CN groups are in ZIF-96, thus leading to high affinity for water. Overall, ZIF-25 and -71 can be classified as hydrophobic, whereas ZIF-96 is hydrophilic. Such classification was confirmed by water adsorption in our recent study.29 More specifically, it was found that water adsorption is vanishingly small in ZIF-25 and -71, but rather high in ZIF-96. The structural analysis above reveals that water has the highest affinity with ZIF-96, followed by ZIF-71 and -25. Once entering the membrane, water is the most strongly bound onto ZIF-96 framework with the least capability to flow; thus, the flux in ZIF-96 is the lowest among the three ZIFs despite the largest da. We infer that the flux in ZIF-25, -71, and -96 is mainly governed by the polarity of functional group, not to follow the increasing trend of da. In the presence of a CH3 group, the most hydrophobic ZIF-25 exhibits the highest flux. With the same reason, this phenomenon of fast water transport was observed in CNTs by both simulations34,35 and experiments.41,42 Hydrogen bonding of water in ZIF-25, -71, and -96 is also examined. Specifically, two geometrical criteria were implemented to define a hydrogen bond: (1) the distance between a donor and an acceptor ≤0.35 nm; (2) the angle of hydrogendonor−acceptor ≤30°.43,44 On average, one water molecule

Figure 7. Autocorrelation function c(t) of hydrogen bonding in ZIF25, -71, and -96 membranes and in bulk phase, respectively.

c(t) values in ZIF-25, -71, and -96 at a given time are larger than in bulk phase, which implies the confinement and framework interaction restrict water motion in the membranes and leads to longer existence of hydrogen bonds. Among the three membranes, c(t) decays the fastest in ZIF-25 and then ZIF-71 and -96, following the decreasing trend of water flux. Obviously, at a higher water flux, fewer hydrogen bonds would remain. The lifetime τHB are 1.7, 1.8, and 2.2 ps in ZIF-25, -71 and -96, respectively. They are longer than 1.4 ps in bulk water but shorter than 3.5 ps in ZIF-8.22 ZIF-25, -71, and -96 possess an aperture larger than that of ZIF-8, resulting in a higher water flux; therefore, the relaxation of hydrogen bonds in ZIF-8 is slower. Compared with the case in a (10, 0) CNT (τHB > 15 ps),36 however, the lifetime in the three ZIFs is considerably shorter. This is because water molecules are packed far more compactly in one-dimensional small CNT than in threedimensional network of ZIFs, and a longer time is required for hydrogen bonds to form and break. It is instructive to investigate the mechanism of water transport through the ZIF membranes. Shown in Figure 8 are the trajectories of randomly selected water molecules passing through three membranes ZIF-25, -96, and -97. As presented above, water transport through ZIF-25 is the fastest due to the large aperture and framework hydrophobicity. Once entering E

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 8. Trajectories of selected water molecules through ZIF-25, -96, and -97 membranes. Each membrane is between the two dotted lines.

Figure 9. Water flux Jw through ZIF-25 membrane as a function of (a) pressure gradient ΔP and (b) inverse temperature (1000/T) at ΔP = 90 MPa. The dashed line in b fits to the Arrhenius equation.

ZIF-25, all the selected water molecules flush through the membrane within 0.5 ns. In ZIF-96 with a slightly larger aperture but polar functional group, a few water molecules pass through the membrane within 1 ns; however, many stay in the membrane for 3−4 ns due to the strong framework affinity, and they jump forward and backward before leaving the membrane. In ZIF-97 with a small aperture, water molecules stay in the membrane frequently jumping around up to 6−7 ns. The jumping motion was commonly observed for water in polyamides tested as RO membranes.46,47 From this analysis, water transport through the ZIF membranes can be considered to follow two modes (flushing and jumping), which depends on the aperture size and polarity of functional group. 3.3. Effects of Pressure and Temperature on Water Transport in ZIF-25. The results in section 3.1 are at ΔP = 60 MPa and 300 K. Figure 9a plots water flux Jw through ZIF-25 as a function of ΔP. Upon increasing ΔP from 30 to 90 MPa, Jw increases in a linear manner. At ΔP = 90 MPa, the net transferred water molecules through a single pore of ZIF-25 is about 43 per ns, much more than that through ZIF-8 (2.1 at ΔP = 90 MPa)22 or (6, 6) CNT (10 at ΔP = 100 MPa).35 The reason is that ZIF-25 has a larger aperture (5.1 Å) than ZIF-8 (3.4 Å) and (6, 6) CNT (4.7 Å). Water transport in a membrane generally follows the Arrhenius equation

⎛ −E ⎞ Jw ∼ exp⎜ a ⎟ ⎝ RT ⎠

function of 1000/T, the Ea through ZIF-25 is estimated to be 6.5 kJ/mol. This value is much lower than that through ZIF-8 (24.4 kJ/mol)22 or FT30 membrane (ranging from 22 to 26 kJ/ mol at 5−40 °C).48 The reason is that ZIF-25 has an aperture larger than that of ZIF-8 and FT30 membrane, and moreover a hydrophobic framework; consequently, the energy barrier is lower for water transport through ZIF-25. Compared with 1.6 kJ/mol in (6, 6), (7, 7), and (8, 8) CNTs with similar aperture size,49 the Ea in ZIF-25 is higher because CNTs are smooth and highly hydrophobic, allowing for very fast water transport with extremely low energy barrier.

4. CONCLUSIONS Molecular simulations have been conducted to investigate water desalination via RO through ZIF-25, -71, -93, -96, and -97 membranes. Due to the presence of various functional groups, the aperture size and polarity differ in the five ZIFs. With a larger aperture size (da = 5.1−5.5 Å), ZIF-25, -71, and -96 possess water flux much higher than ZIF-93 and -97 (da = 3.5− 3.7 Å). From the analysis of radial distribution functions, the framework affinity for water is found to increase as ZIF-25 < -71 < -96. Therefore, water is the most loosely bound onto ZIF25 but strongly onto ZIF-96; water flux decreases as ZIF-25 > -71 > -96, following the opposite trend of framework affinity. The most hydrophobic ZIF-25 shows the highest water flux, a phenomenon also observed in highly hydrophobic carbon nanotubes. The ZIFs are predicted to have water permeance/ permeability significantly higher than MFI zeolite, commercial seawater RO, brackish RO, and high-flux RO membranes. Depending on the aperture size and polarity of functional

(4)

where Ea, R, and T are activation energy, gas constant, and temperature, respectively. From Figure 9b showing Jw as a F

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir group, water transport in the ZIFs follows two modes: flushing and jumping. The former is observed in ZIF-25, while the latter frequently occurs in ZIF-96 and -97. Because of the confinement and framework interaction, the lifetime of hydrogen bonding in ZIF-25, -71, and -96 are 1.7, 1.8, and 2.2 ps, respectively, all longer than 1.4 ps in bulk phase. Finally, it is found that water flux in ZIF-25 scales linearly with pressure gradient and exhibits an Arrhenius-type relationship with temperature. The activation energy for water transport in ZIF-25 is estimated to be 6.5 kJ/mol, lower than in ZIF-8 and polyamide membranes but higher than in CNTs. This simulation study suggests that ZIF-25 might be an intriguing RO membrane for water desalination. Nevertheless, we should note several limitations associated with the simulations: (1) The ZIFs were assumed to be rigid and chemically stable in water. If the framework flexibility is taken into account, salt rejection as well as water flux may be affected. Moreover, their stability needs further testing. (2) In order to achieve observable RO process within a nanosecond time scale, the RO pressures applied were much higher than in practical cases. (3) The ZIF membranes modeled were very thin and cannot be straightforwardly compared with realistic membranes. Despite these limitations, this study provides the important molecular-level understanding of water structural and dynamic properties in various ZIFs, unravels the significant role of functional groups, and is helpful to the rational design of new porous materials from the bottom-up toward highperformance water desalination.

■

(4) Elimelech, M.; Phillip, W. A. The future of seawater desalination: energy, technology, and the environment. Science 2011, 333, 712−717. (5) Khawaji, A. D.; Kutubkhanah, I. K.; Wie, J.-M. Advances in seawater desalination technologies. Desalination 2008, 221, 47−69. (6) International Desalination Association. IDA Desalination Yearbook 2011−2012; Media Analytic Ltd.: Oxford, UK, 2011. (7) Pendergast, M. T. M.; Hoek, E. M. V. A review of water treatment membrane nanotechnologies. Energy Environ. Sci. 2011, 4, 1946−1971. (8) Duke, M. C.; O’Brien-Abraham, J.; Milne, N.; Zhu, B.; Lin, J. Y. S.; Diniz da Costa, J. C. Seawater desalination performance of MFItype membranes made by secondary growth. Sep. Purif. Technol. 2009, 68, 343−350. (9) Zhu, B.; Hong, Z.; Milne, N.; Doherty, C. M.; Zou, L.; Lin, Y. S.; Hill, A. J.; Gu, X.; Duke, M. Desalination of seawater ion complexes by MFI-type zeolite membranes: Temperature and long term stability. J. Membr. Sci. 2014, 453, 126−135. (10) Das, R.; Ali, M. E.; Hamid, S. B. A.; Ramakrishna, S.; Chowdhury, Z. Z. Carbon nanotube membranes for water purification: A bright future in water desalination. Desalination 2014, 336, 97−109. (11) Huang, L.; Zhang, M.; Li, C.; Shi, G. Graphene-based membranes for molecular separation. J. Phys. Chem. Lett. 2015, 6, 2806−2815. (12) Aghigh, A.; Alizadeh, V.; Wong, H. Y.; Islam, M. S.; Amin, N.; Zaman, M. Recent advances in utilization of graphene for filtration and desalin of water: A review. Desalination 2015, 365, 389−397. (13) Eddaoudi, M.; Kim, J.; Rosi, N.; Vodak, D.; Wachter, J.; O’Keefe, M.; Yaghi, O. M. Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science 2002, 295, 469−472. (14) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The chemistry and applications of metal-organic frameworks. Science 2013, 341, 1230444. (15) Li, J. R.; Kuppler, R. J.; Zhou, H. C. Selective gas adsorption and separation in metal-organic frameworks. Chem. Soc. Rev. 2009, 38, 1477−1504. (16) Keskin, S.; Liu, J.; Rankin, R. B.; Johnson, J. K.; Sholl, D. S. Progress, opportunities, and challenges for applying atomically detailed modeling to molecular adsorption and transport in metal−organic framework materials. Ind. Eng. Chem. Res. 2009, 48, 2355−2371. (17) Jiang, J. W.; Babarao, R.; Hu, Z. Q. Molecular simulations for energy, environmental and pharmaceutical applications of nanoporous materials: from zeolites, metal-organic frameworks to protein crystals. Chem. Soc. Rev. 2011, 40, 3599−3612. (18) Suh, M. P.; Park, H. J.; Prasad, T. K.; Lim, D. W. Hydrogen storage in metal−organic frameworks. Chem. Rev. 2012, 112, 782−835. (19) Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.; Bae, T. H.; Long, J. R. Carbon dioxide capture in metal−organic frameworks. Chem. Rev. 2012, 112, 724− 781. (20) Getman, R. B.; Bae, Y. S.; Wilmer, C. E.; Snurr, R. Q. Review and analysis of molecular simulations of methane, hydrogen and acetylene storage in metal−organic frameworks. Chem. Rev. 2012, 112, 703−723. (21) Yang, Q. Y.; Liu, D. H.; Zhong, C. L.; Li, J. R. Development of computational methodologies for metal−organic frameworks and their application in gas separations. Chem. Rev. 2013, 113, 8261−8323. (22) Hu, Z. Q.; Chen, Y. F.; Jiang, J. W. Zeolitic imidazolate framework-8 as a reverse osmosis membrane for water desalination: Insight from molecular simulation. J. Chem. Phys. 2011, 134, 134705. (23) Duan, J.; Pan, Y.; Pacheco, F.; Litwiller, E.; Lai, Z.; Pinnau, I. High-performance polyamidethin-film-nanocomposite reverse osmosis membranes containing hydrophobic ZIF-8. J. Membr. Sci. 2015, 476, 303−310. (24) Willems, T. F.; Rycroft, C. H.; Kazi, M.; Meza, J. C.; Haranczyk, M. Algorithms and tools for high-throughput geometry-based analysis of crystalline porous materials. Microporous Mesoporous Mater. 2012, 149, 134−141.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b03593. Fragmental clusters of ZIFs, force field parameters and atomic charges of ZIFs, number distributions of ions at different times, number of Na+ ions crossing ZIF membranes, and radial distribution functions between water molecules in ZIF membranes (PDF) A movie showing the RO process in the ZIF-96 membrane (MPG)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to the A*star of Singapore (R-279-000-431305) and the National University of Singapore for financial support.

■

REFERENCES

(1) Schiermeier, Q. Water: purification with a pinch of salt. Nature 2008, 452, 260−261. (2) Shannon, M. A.; Bohn, P. W.; Elimelech, M.; Georgiadis, J. G.; Marinas, B. J.; Mayes, A. M. Science and technology for water purification in the coming decades. Nature 2008, 452, 301−310. (3) Service, R. F. Desalination freshens up. Science 2006, 313, 1088− 1090. G

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (25) Morris, W.; Leung, B.; Furukawa, H.; Yaghi, O. K.; He, N.; Hayashi, H.; Houndonougbo, Y.; Asta, M.; Laird, B. B.; Yaghi, O. M. A combined experimental−computational investigation of carbon dioxide capture in a series of isoreticular zeolitic imidazolate frameworks. J. Am. Chem. Soc. 2010, 132, 11006−11008. (26) Banerjee, R.; Phan, A.; Wang, B.; Knobler, C.; Furukawa, H.; O’Keeffe, M.; Yaghi, O. M. High-throughput synthesis of zeolitic imidazolate frameworks and application to CO2 capture. Science 2008, 319, 939−943. (27) Mayo, S. L.; Olafson, B. D.; Goddard, W. A. Dreiding - a generic force-field for molecular simulations. J. Phys. Chem. 1990, 94, 8897− 8909. (28) Zhang, K.; Zhang, L.; Jiang, J. W. Adsorption of C1−C4 alcohols in zeolitic imidazolate framework-8: Effects of force fields, atomic charges, and framework flexibility. J. Phys. Chem. C 2013, 117, 25628−25635. (29) Zhang, K.; Nalaparaju, A.; Chen, Y.; Jiang, J. W. Biofuel purification in zeolitic imidazolate frameworks: The significant role of functional groups. Phys. Chem. Chem. Phys. 2014, 16, 9643−9655. (30) Zhang, K.; Gupta, K. M.; Chen, Y. F.; Jiang, J. W. Biofuel purification in GME zeolitic−imidazolate frameworks. AIChE J. 2015, 61, 2763−2775. (31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 09, Revision D.01 ed.; Gaussian, Inc.: Wallingford, CT, 2009. (32) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 1983, 79, 926−935. (33) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. A 2nd generation force-field for the simulation of proteins, nucleic-acids, and organic-molecules. J. Am. Chem. Soc. 1995, 117, 5179−5197. (34) Hummer, G.; Rasaiah, J. C.; Noworyta, J. P. Water conduction through the hydrophobic channel of a carbon nanotube. Nature 2001, 414, 188−190. (35) Suk, M. E.; Raghunathan, A. V.; Aluru, N. R. Fast reverse osmosis using boron nitride and carbon nanotubes. Appl. Phys. Lett. 2008, 92, 133120. (36) Suk, M. E.; Aluru, N. R. Water transport through ultrathin graphene. J. Phys. Chem. Lett. 2010, 1, 1590−1594. (37) Cohen-Tanugi, D.; Grossman, J. C. Water desalination across nanoporous graphene. Nano Lett. 2012, 12, 3602−3608. (38) Kou, J.; Zhou, X.; Lu, H.; Wu, F.; Fan, J. Graphyne as the membrane for water desalination. Nanoscale 2014, 6, 1865−1870. (39) Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A. E.; Berendsen, H. J. C. Gromacs: Fast, flexible, and free. J. Comput. Chem. 2005, 26, 1701−1718. (40) Liu, X.; Demir, N. K.; Wu, Z.; Li, K. Highly water-stable zirconium metal−organic framework uio-66 membranes supported on alumina hollow fibers for desalination. J. Am. Chem. Soc. 2015, 137, 6999−7002. (41) Majumder, M.; Chopra, N.; Andrews, R.; Hinds, B. Nanoscale hydrodynamics enhanced flow in carbon nanotubes. Nature 2005, 438, 44. (42) Holt, J. K.; Park, H. G.; Wang, Y.; Stadermann, M.; Artyukhin, A. B.; Grigoropoulos, C. P.; Noy, A.; Bakajin, O. Fast mass transport through sub-2-nanometer carbon nanotubes. Science 2006, 312, 1034− 1037.

(43) Ferrario, M.; Haughney, M.; McDonald, I. R.; Klein, M. L. Moleculardynamics simulation of aqueous mixtures: Methanol, acetone, and ammonia. J. Chem. Phys. 1990, 93, 5156−5166. (44) Luzar, A.; Chandler, D. Hydrogen-bond kinetic in liquid water. Nature 1996, 379, 55−57. (45) Xu, H. F.; Berne, B. J. Hydrogen-bond kinetics in the solvation shell of a polypeptide. J. Phys. Chem. B 2001, 105, 11929−11932. (46) Kotelyanskii, M. J.; Wagner, N. J.; Paulaitis, M. E. Molecular dynamics simulation study of the mechanisms of water diffusion in a hydrated, amorphous polyamide. Comput. Theor. Polym. Sci. 1999, 9, 301−306. (47) Sharma, V. K.; Singh, P. S.; Gautam, S.; Maheshwari, P.; Dutta, D.; Mukhopadhyay, R. Dynamics of water sorbed in reverse osmosis polyamide membrane. J. Membr. Sci. 2009, 326, 667−671. (48) Mehdizadeh, H.; Dickson, J. M.; Eriksson, P. K. Temperature effects on the performance of thin-film composite, aromatic polyamide membranes. Ind. Eng. Chem. Res. 1989, 28, 814−824. (49) Corry, B. Designing carbon nanotube membranes for efficient water desalination. J. Phys. Chem. B 2008, 112, 1427−1434.

H

DOI: 10.1021/acs.langmuir.5b03593 Langmuir XXXX, XXX, XXX−XXX