Li+ Separation: The

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Mg2+-channel-inspired nanopores for Mg2+/Li+ separation: the effect of coordination on the ionic hydration microstructures Yudan Zhu, Yang Ruan, Yumeng Zhang, Yaojia Chen, Xiaohua Lu, and Linghong Lu Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b01249 • Publication Date (Web): 13 Aug 2017 Downloaded from http://pubs.acs.org on August 18, 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.

Langmuir 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 26

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

Langmuir

Mg2+-channel-inspired nanopores for Mg2+/Li+ separation: the effect of coordination on the ionic hydration microstructures Yudan Zhu*, Yang Ruan, Yumeng Zhang, Yaojia Chen, Xiaohua Lu, Linghong Lu College of Chemical Engineering, State Key Laboratory of Materials-oriented Chemical Engineering, Nanjing Tech University, Nanjing 210009, P.R. China

* Author to whom correspondence should be addressed: Yudan Zhu, Email address:[email protected]

ACS Paragon Plus Environment

Langmuir

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

Abstract 2+

The separation behaviors of Mg

and Li+ were investigated using molecular dynamics.

Two functionalized graphene nanopore models (i.e., co_5 and coo_5) inspired by the characteristic structural features of Mg2+ channels were used. Both nanopores exhibited a higher preference to Mg2+ than to Li+, and the selectivity ratios were higher for coo_5 than for co_5 under all the studied transmembrane voltages. An evaluation of the effect of coordination on the ionic hydration microstructures for both nanopores showed that the positioning of the modified groups could better fit a hydrated Mg2+ than a hydrated Li+, as if Mg2+ was not dehydrated according to hydrogen bond analysis of the ionic hydration shells. This condition led to a lower resistance for Mg2+ than for Li+ when traveling through the nanopores. Moreover, a distinct increase in hydrogen bonds occurred with coo_5 compared with co_5 for hydrated Li+, which made it more difficult for Li+ to pass through coo_5. Thus, a higher Mg2+/Li+ selectivity was found in for coo_5 than for co_5. These findings provide some design principles for developing artificial Mg2+ channels, which have potential applications as Mg2+ sensors and novel devices for Mg2+/Li+ separation.


Page 2 of 26

Page 3 of 26

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

Langmuir

INTRODUCTION Lithium-ion batteries are a promising option for powering electric devices and vehicles. Salt lake brine has become the main source of lithium because of the increasing demand of lithium resources.1 Currently, conventional Mg2+/Li+ separation methods are widely used.2, 3 However, as the most interfering element, Mg2+ has similar ionic properties to Li+ that accounts for the inevitable Mg2+ residue during lithium extraction, particularly for the high ratio Mg2+/Li+ salt lake. Membrane separation for Mg2+/Li+ shows a promising future despite facing several challenges.4 Given that the mechanism of ion transport at the nanoscale is not fully understood,5 studying the difference in the ionic transport resistances of Mg2+ and Li+ in nanopores is essential. Nature provides smart solutions to achieve efficient functions; thus, learning from characteristic structural features in nature has long been a source of inspiration for academics.6 Synthetic ion channels have become a hot topic both in academia and industry7-9 considering that biological ion channels exhibit efficient functions that allow for the rapid transport of certain types of ions across cell membranes. K+ channels have been one of the most frequently studied channels since 2003 when the Nobel Prize in Chemistry was awarded to the discoverer of KcsA.10 In the classical mechanism, the positioning of the carbonyl groups in the KcsA filter region can precisely fit a hydrated K+ but not a hydrated Na+. As a result, the cell allows K+ transport through the lipid while excluding Na+. The pioneering research on the characterization of K+ channels by MacKinnon11 not only exhibited the elaborate structure of K+ channels but also provided useful and fundamental principles for the further development of biomimetic K+ channels. Up to now, as far as the mechanism of biological K+ channels is considered, new findings are unceasingly reported.12, 13 Free-energy analysis based on potentials of mean force (PMF) is indispensable for investigating the selectivity mechanism in biological K+ channels14 and biomimetic K+ nanopores.15,

16

Meanwhile, there is considerable evidence17-20 indicating that the ionic

hydration microstructure variation influenced by the specific channel (pore) structure should not be ignored. He et al.15 and Kang et al.16 preformed both the PMF and ionic hydration


Langmuir

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 26

microstructure analyses in their investigations on the K+/Na+ separation behaviors of biomimetic K+ nanopores. Their results indicated that K+ can be well coordinated by oxygen atoms in the modified groups; this reduces the energy barrier for K+ with respect to Na+, leading to the separation of K+/Na+. Recently, the role of dehydration in ion separation was emphasized by Sahu et al.21 They noted that dehydration not only provides ion-dependent free-energy barriers but is also predicted to give rise to selectivity. These findings provide useful guidelines for investigating the selectivity of other ion channels or nanopores. Mg2+ is the second most abundant biological cation after K+. The excellent selectivity of Mg2+ channels allows for the Mg2+ transport into the cell through the lipid, while excluding other monovalent cations.22,

23

In contrast to the studies on K+, related investigations on

biomimetic Mg2+ nanopores are not common. With inspiration from the characteristic sizes of the selectivity filter of biological Mg2+ channels, we previously adopted the molecular simulation to investigate Mg2+/Li+ separation and found that in pristine graphene-based nanopores, the difference in the ionic dehydration of the second hydration shells of Mg2+ and Li+ is the main reason.24 Recent findings showed that two important binding sites exist in the Thermotoga maritime (TmCorA) Mg2+ channel to recognize Mg2+ (i.e., one is a selectivity filter formed by five (Gly-Met-Asn) GMN motifs in the extracellular loop, and the other is a divalent cation sensor (DCS) site at the N-terminal cytoplasmic domain).25 As shown in Figure 1, the carbonyl and negatively charged carboxylate groups are separately, major components in both sites. We speculated that the specific structures formed by the modified groups may have a unique effect on the ionic hydration microstructures of different ions, which can be associated with ionic selectivity. The present work aims to verify this prediction and provide molecular information for designing biomimetic Mg2+ channels. With the advancement of experimental techniques, many novel two-dimensionalmaterial-based nanopores (e.g., graphene oxide membranes,26,

27

graphene-oxide-based

membranes functionalized with a peptide motif,28 crown ethers configurations in graphene,29 and single-layer MoS2 nanopores30) have been developed. In this study, two simple and functionalized graphene nanopores were adopted as models, which are similar to models that


Page 5 of 26

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

Langmuir

were used in previous simulation investigations on biomimetic nanopores.15, 16, 31-34 Molecular dynamics calculations were performed to investigate the separation behaviors of Mg2+ and Li+ passing through the nanopore models under various electric field intensities. Bioinspired by the characteristic pentamer structural features of metal-ion binding sites in Mg2+ channels, we focused on evaluating the effect of carbonyl and carboxylate groups on Mg2+/Li+ selectivity. To better unravel the underlying mechanism of selectivity in these two nanopores, the results of PMF and detailed ionic hydration microstructure analyses are discussed. SIMULATION MODELS AND METHODS Figure 1 shows the two graphene nanopore models. Each nanopore was obtained by removing 24 carbon atoms from the central region of graphene. The diameter of the pristine graphene nanopores was 1.024 nm. Inspired by biological Mg2+ channels,25 five carbonyl groups were attached to the nanopore (labeled as “co_5”) to mimic the characteristic pentamer structure of the selectivity filter in Mg2+ channels. For comparison, the nanopore modified with five negatively charged carboxylate groups (labeled as “coo_5”) was also investigated, considering its key role in the DCS sites. The graphene sheet was in the center of a three-directional periodic simulation box (4.18 ×4.25 ×5.10 nm3), parallel to the xy-plane. For each case, the simulation box was filled with 0.25 M MgCl2 and LiCl mixed solution. Counter ions were added accordingly to maintain electro-neutrality for the simulation case. All the details of the simulated systems are listed in Table S1.

Figure 1. (a) Structure of Thermotoga maritime (TmCorA) Mg2+ channels. (b) Above: The lateral view of the simulation box that consists of a modified graphene nanopore in the center and reservoirs on both sides


Langmuir

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

of the nanopore. Below: The top view of the co_5 and coo_5 nanopores modified by five carbonyl groups and five negatively charged carboxylate groups, respectively.

Water molecules were described by the TIP3P model.35 During the simulation, the carbon atoms on the edges of the graphene sheet were fixed at their initial positions, while others and their connected modified groups were allowed to relax.15, 16 An OPLS/AA force field36 was used to describe the graphene carbon atoms and ions. The electronic continuum correction method37, 38 was implemented for the charge of Li+. The ionic scaling factor for the water solvent was usually 0.75. With the exception of the electronic continuum correction of charge, other parameters were obtained from the OPLS/AA force field. For the modified groups, the partial charges and parameters for the carbonyl and negatively charged carboxylate groups were obtained from ASN and ASP amino acids, respectively. Each carboxylate group carried a net charge of -1 e. He et al.15 adopted a similar method to study the system of a modified graphene nanopore. The Lennard–Jones interactions between different particles used the Lorentz−Berthelot combination rule.39 The short-range interactions utilized a cutoff of 1.0 nm, whereas the long-range electrostatic interactions adopted the particle-mesh Ewald (PME) method40 with a cutoff for a real space of 1.0 nm. Gromacs41 was used to conduct molecular dynamic (MD) simulations, and the Visual Molecular Dynamics (VMD) package42 was used for visualization. Each simulation system was first equilibrated for 5 ns in the NPT ensemble after energy minimization and then ran for 105 ns in the NVT ensemble with a 2 fs time step. All simulation data were saved every 1 ps for additional analysis. The pressure and temperature were maintained at 300 K and 1 bar, respectively, using the V-rescale method43 and the Parrinello−Rahman coupling scheme.44, 45 In the present study, the five electric field intensities were 0.2, 0.4, 0.6, 0.8 and 1.0 V/nm and the corresponding transmembrane voltages were 1.02, 2.04, 3.06, 4.08 and 5.10 V. RESULTS AND DISCUSSION Ionic flux and Mg2+/Li+ selectivity ratio Figure 2 shows the two kinds of ionic fluxes (Mg2+ and Li+) of the different modified nanopores under different electric field intensities and their corresponding Mg2+/Li+ selectivity


Page 6 of 26

Page 7 of 26

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

Langmuir

ratios. The ionic flux was counted as the number of a certain type of ion passing through the nanopore under an applied electric field. As illustrated in Figure 2, the flux of Mg2+ was higher than that of Li+ under all the studied transmembrane voltages for both co_5 and coo_5 nanopores, indicating that both modified pores had a preference to Mg2+ over Li+. The Mg2+ flux increased as the electric field intensity increased, whereas no discernible dependence was observed for the Li+ flux and the Mg2+/Li+ selectivity ratio on the electric field intensity. The Mg2+/Li+ selectivity ratios for coo_5 were higher than those for co_5. The highest selectivity ratios of Mg2+/Li+ were 3.56 and 4.73 for co_5 and coo_5, respectively, which are high compared with the pristine nanopore of the previous work.24

s 2+

+

Figure 2. Ionic fluxes and Mg /Li selectivity ratio of passing through (a) co_5 and (b) coo_5 graphene-based nanopores in 100 ns MD simulations under different electric field intensities.

Furthermore, we calculated the PMF for Mg2+ and Li+ along the z-direction in the absence of an electric field, as shown in Figure S1. For both nanopores, the PMF profiles of Mg2+ were less than zero, whereas the ones of Li+ were greater than zero. The results indicate that Mg2+ tended to enter the nanopores, but Li+ encountered barriers at the pore entrances. This phenomenon can explain why both nanopores exhibited a preference to Mg2+ higher than that to Li+ in terms of free energy. The evident difference in the PMF profiles of co_5 and coo_5 could indicate that different modified groups had different influences on the ions. The highest free energy barriers encountered by Li+ passing through the nanopore were higher for coo_5 than for co_5, which suggested that the transport resistance for Li+ was higher for coo_5 than for co_5. This phenomenon could explain to some extent the reason why the selectivity was higher in coo_5 than in co_5. In general, the PMF results could explain the selectivity results based on free energy.


Langmuir

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

Beckstein et al.46 found that ionic hydration cannot be ignored when considering ions passing through a nanopore. We deduced that, at the molecular scale, the different roles of the modified groups on the structure of the ionic hydration shells might result in the different Mg2+/Li+ selectivity ratios in both kinds of nanopore. Thus, in the following section, we focus on the changes in the microstructural properties of the ionic hydration shell when the ions were traveling through the nanopore to further elucidate the ion separation mechanism at the molecular scale. Ionic hydration microstructure analysis Ions were counted when they were within the imaginary cylinder with its axis perpendicular to the graphene sheet and with its diameter equal to the pore diameter. We divided the ions into two regions to reflect the nanopore confinement effect: (i) at the entrance of the nanopore (-0.25 nm < z < 0.25 nm, where z = 0 represents the pore position) and (ii) in the bulk solution (z < -0.6 nm, z > 0.6 nm). Because of the strong ionic hydration of Mg2+ and Li+, two hydration shells for the ions were considered, as illustrated in Figure 3. According to the method described in our previous work24, we obtained the ionic hydration radii for different shells. The first hydration shell radii were approximately 0.268 and 0.244 nm for Li+ and Mg2+, respectively, and the second hydration shell radii were 0.556 and 0.516 nm, respectively.

Figure 3. Schematic image of the ionic hydration microstructure (green: the first hydration shell, gray: the second hydration shell)


Page 8 of 26

Page 9 of 26

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

Langmuir

Hydration number distribution in the second ionic hydration shell In our previous work,24 we found that the dehydration of the second hydration shell of ions governs the selectivity of Mg2+/Li+ for pristine graphene nanopores. Thus, in the present work, we first focused on the hydration number distributions of ions in the second hydration shell at the entrance of a nanopore and in the bulk solution under different electric field intensities, as shown in Figure 4. For all electric field intensities, the distribution profiles were nearly identical for a particular nanopore, indicating that the electric field intensity had a slight effect on the hydration number distribution. For co_5 (see Figure 4a and b), the distribution peaks of Mg2+ appeared at 10-11 and the bulk one was at approximately 15. The distribution peaks of Li+ were concentrated on 15-16, and the bulk one was approximately 20. The clear decrease in the hydration number indicated that dehydration occurred in the second hydration shell when ions were passing through the nanopore. The dehydration numbers were nearly the same for Li+ and Mg2+. For coo_5, the distribution profiles were similar to those for co_5, suggesting that the degree of dehydration of the two kinds of modified groups was nearly equal. Moreover, we analyzed the hydration number distributions of ions in the first hydration shell, as shown in Figure S4. Similar profile peaks between the nanopore entrance and the bulk demonstrated that dehydration primarily occurred in the second hydration shell for co_5 and coo_5, although slight differences exist in the Li+ profile under different electric field intensities. Although the dehydration numbers of Mg2+ and Li+ for both modified nanopores had no clear difference, the selectivity values of Mg2+/Li+ were different for the two modified nanopores. To account for these phenomena, we turn to a previous investigation on the biological K+ channel. MacKinnon et al.11 found that the carbonyl oxygen atoms on the selectivity filter of KcsA can perfectly substitute the role of water molecules in the K+ hydration shell but fail to substitute its role in the Na+ hydration shell, which leads to a remarkable difference in preference for K+ and Na+. These findings suggest that whether the modified groups can substitute the role of water molecules in the ionic hydration shell or not


Langmuir

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

determines the selectivity, as well as the dehydration number. Thus, we further evaluated the effect of the modified groups on the ionic hydration microstructure.

Figure 4. Hydration number distributions of Mg2+ and Li+ in the second hydration shell at the entrance of a nanopore and in the bulk solution under different electric field intensities. (a) co_5: Mg2+, (b) co_5: Li+, (c) coo_5: Mg2+, and(d) coo_5: Li+.

The effects of modified groups on the ionic coordination number and water molecule orientation in hydration shells In this section, we first analyze the influence of modified groups on the coordination number of ionic hydration shell and examine whether the coordination number was saturated (Figures 5 and 6). Second, we focus on the influence of the modified groups on the water molecule orientations in the ionic hydration shells (Figure 7). Mg2+ and Li+ had evident dehydration in co_5 and coo_5, which implied that both kinds of modified groups might enter the second hydration shells. Thus, we investigated the coordination number distributions of Mg2+ and Li+ in the second hydration shell at the entrance of the nanopores under different electric field intensities (see Figure 5). The coordination numbers were obtained by counting both the number of water molecules and oxygen atoms of the modified groups in the second ionic hydration shells. As shown in Figure


Page 10 of 26

Page 11 of 26

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

Langmuir

5a, the profile peaks in the coordination number distributions of Mg2+ at the pore entrance of co_5 were nearly consistent with their bulk counterparts and were concentrated on 15. However, the distributions at the pore entrance were much sharper. In contrast to Mg2+, Li+ showed a clear shift in coordination number distributions, and the difference was approximately 2 (Figure 5b). These observations indicated that the modified groups in co_5 could replace the removed water molecules so that the coordination numbers in the second ionic hydration shell of Mg2+ at the pore entrance were similar to the bulk saturated one, making the hydration microstructures of Mg2+ stable when passing through the nanopore. However, the hydrated Li+ was unsaturated at the pore entrance, although the modified groups in co_5 replaced some water molecules. The effect of the modified groups in coo_5 on the coordination numbers of Mg2+ and Li+ in the second ionic hydration shell was similar to that in co_5, except that coo_5 had a slight difference in the coordination numbers of Mg2+ at the pore entrance and in the bulk.

Figure 5. Coordination number distributions of Mg2+ and Li+ in the second hydration shell at the entrance of the nanopores under different electric field intensities. (a) co_5: Mg2+, (b) co_5: Li+, (c) coo_5: Mg2+ and (d) coo_5: Li+. The ionic hydration distributions in their bulk counterparts are given for comparison.

We further investigated the number distributions of oxygen atoms in the modified groups entering the second hydration shells of Mg2+ and Li+. As illustrated in Figure 6a and c, the


Langmuir

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

most likely number of oxygen atoms for Mg2+ was 5 for both co_5 and coo_5. On the other hand, boarder distributions and approximately of 2–5 and 2–4 oxygen atoms exist for co_5 and coo_5, respectively, for Li+. Given that the dehydration numbers of Mg2+ and Li+ in their second hydration shells were concentrated at 4–5 (in Figure 4a and c), Figure 6 affirms that the oxygen atoms in the modified groups replaced the removed water in the second hydration shell of Mg2+ and retained the saturated coordination number in the ionic second hydration shell of Mg2+. Meanwhile, the coordination number in the Li+ second hydration shells was unsaturated. In their investigation of the selectivity filter in biological Mg2+ channel, Kitjaruwankul et al.25 revealed that five GMN motifs enhanced the affinity to an ion by acting as a coordination ligand in the second ionic hydration shell. In this work, Mg2+ passing through both co_5 and coo_5 possessed similar characteristic features that were similar to the biological Mg2+ channel, which suggested that the pentameric pattern was essential for the Mg2+ channel selectivity filter. The major difference in the coordination saturation of the ionic second hydration shells of Mg2+ and Li+ was the primary cause for the preference of Mg2+ to Li+ in co_5 and coo_5.


Page 12 of 26

Page 13 of 26

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

Langmuir

Figure 6. Number distributions of the oxygen atoms in modified groups entering the second hydration shells of Mg2+ and Li+ under different electric field intensities. (a) co_5: Mg2+, (b) co_5: Li+, (c) coo_5: Mg2+, and (d) coo_5: Li+.

Moreover, the different degrees of coordination saturation between Mg2+ and Li+ in their second hydration shells might be relevant to the ionic first hydration shell variations when the ions were passing through the modified nanopores. In our previous studies,24 Mg2+ and Li+ have different preferential ionic transport pathways. Mg2+ prefers to pass through the central region of the nanopore, whereas Li+ prefers to pass near the pore edge. In the present study, Figure S5 reveals similar preferences. As shown in Figure S5, 1-2 oxygen atoms in the modified groups could enter the Li+ first hydration shell, but none could for Mg2+. This observation indicated that the hydration of Li+ could be more easily disturbed by modified groups. For Mg2+, the modified groups did not substitute these water molecules because of the strong electrostatic interactions between Mg2+ and the water molecules of the first hydration shells. As a result, Mg2+ mostly passed through the nanopore central region. In biological Mg2+ channels, the first hydration shell structure of Mg2+ in the filter was also proven to not be considerably different from their bulk counterpart.25, 47 Furthermore, we evaluated the influence of the modified groups on the orientation of the water molecules in the ionic hydration shells. We adopted a parameter named the hydration factor (F)5, 48 to quantify the orientation order of the ionic hydration shell as defined in Eqs. 1a and b. For a cation, F is defined as the ratio of the number of water molecules in the hydration shell with cos θ less than −0.72 to the amount of water molecules in the hydration shell, where θ is the angle between the cation and the dipole moment of the water molecule. F was counted when the modified groups entered the ionic hydration shell. Based on the definition, the closer the value of F to 1, the higher the order of the hydration shell. Conversely, a value of F that is closer to 0.14 indicates a random orientation distribution of water molecules. The average hydration factors of Mg2+ and Li+ at the entrance of the nanopores and in the bulk solution under different electric field intensities are shown in Figure 7. shell N −first 1< cosθ

Li+ Separation: The

Recommend Documents