Deep dehydrated layered membrane for selective and fast ion separation

Deep dehydrated layered membrane for selective and fast ion separation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Deep dehydrated layered membrane for selective and fast ion separation Jun Gao, Zhaoyu Ma, Jilong Fan, Chenguang Zhu, Hongfei Gao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7726847/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Two-dimensional layered membranes show great promise for separation applications but long suffered from the swelling problem even when intercalated with binding agents such as cations. Swelling is initiated by hydration forces yet restrained by van der Waals attraction and interlayer-ion-bridged electrostatic interactions—both of which increase sharply with the decrease of interaction distance. We therefore demonstrate a deep dehydration strategy to tackle the swelling problem. The deep dehydration at high temperature almost completely removes interlayer water molecules, eliminating hydration-induced repulsion and thus greatly compresses the interlayer spacings. This in turn substantially enhances interlayer attraction and makes rehydration energetically unfavorable. Our deep-dehydrated Cs + -intercalated vermiculite membrane maintains a narrow interlayer spacing of < 0.3 nm in water for over one month, with a swelling ratio of only 4%. These stable nanochannels achieve Li + /Mg 2+ selectivity of 148 and Li + permeability of 0.65 mol m − 2 h − 1 , ranking among the best reported performances, and enable production of industrial-grade lithium carbonate via electrodialysis–precipitation. Our results highlight the critical influence of hydration/dehydration history on layered membrane swelling and performance. Physical sciences/Materials science Physical sciences/Nanoscience and technology Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Two-dimensional (2D) layered membranes, obtained by assembling the 2D nanosheets exfoliated from their bulk form, are regarded highly promising materials for molecular separation due to their facile fabrication method, high scalability, as well as the high structural and chemical tunability 1 – 6 . However, their practical application is severely hindered by their compromised selectivity upon swelling in solvents, particularly in water. For example, neat graphene oxide and vermiculite layered membranes easily disintegrate in water 7 – 9 , and neat MXene membranes swell by over 20% in water 10 – 12 . To address this problem, we should first revisit the mechanism of swelling. For layered materials, it is known that the attractive Coulomb interaction mediated by the intercalated counter-ions and the Van der Waals interaction between nanosheet layers prevent swelling while the hydration of the intercalated counter-ions initiates swelling 13 . Consequently, the most common practice to mitigate swelling is to cross-link the membranes with counter-ions, especially with multivalent ions which has stronger Coulomb interaction 11 , 14 . Yet, such method experienced mixed success because multivalent ions also have larger hydration enthalpy and still cause a large swelling ratio. Monovalent ions, on the other hand, exhibit weak Coulomb interaction and can be easily exchanged out during the separation process, losing the cross-linking capability. We recall that both Coulomb interaction and Van der Waals interaction steeply increase with the decrease of the interaction distance, whereas the hydration interaction does not 15 – 17 . This suggests that if we squeeze the interlayer spacing to an extreme extent, the attractive interaction could overtop the hydration interaction and thus prevent the swelling. It is of course difficult to mechanically squeeze the spacing, since in the atomic scale, the interlayer pressure is up to the GPa scale 18 . We here present a simple but effective method to circumvent this thorny issue. We dehydrate the layered membrane at high temperature (> 500°C), a process we term deep dehydration, to remove the free and hydration water molecules in the interlayer spacings. In this way, the hydration interaction is eliminated, and the attractive interactions naturally compress the membrane, forcing the interlayer spacing size crossing the anti-swelling threshold value. We demonstrated the method with alkaline-ion-intercalated vermiculite membranes (VM). The deep dehydration resulted in an extremely narrow d -spacing (10.5 ~ 11.5 Å, representing a free spacing of 1.5 ~ 2.5 Å after subtracting the theoretical monolayer thickness of 0.9 nm). Density functional theory (DFT) calculations reveal that during the compression of the interlayer channel, the intercalated ions were tightly trapped into the minimum-energy positions, hindering their rehydration. More importantly, the extremely narrow interlayer channel greatly amplifies the binding between the intercalated alkaline ions and the vermiculite, rendering the rehydration energetically unfavorable. As a result, the membrane only swells by 0.97 Å (intercalated by K + ) or 0.45 Å (intercalated by Cs + 19,20 , which has lower hydration energy than K + ) even after immersed in pure water for 30 days, whereas the membranes intercalated by the same alkaline ions without deep dehydration quickly swell in water in a few hours. Thanks to the stable and extremely narrow interlayer free spacing, the deep dehydrated Cs + -intercalated membrane is able to separate Li + and Mg 2+ with high selectivity (148), which is crucial but challenging for salt lake lithium extraction 21 – 25 . Interestingly, the permeability of the Li + also reached a high value of 0.65 mol m − 2 h − 1 despite the confinement. We attribute to the fact that the intercalation ions filled the minimum-energy traps, thereby preventing the Li + ions from being trapped and reducing their transport resistance. Results Fabrication and characterization of the deep-dehydrated membranes Figure 1 a illustrates the fabrication process based on ion intercalation and deep dehydration. Vermiculite nanosheets (characterized in Fig. S1 -S3) exfoliated from the bulk crystals, were dispersed in water and vacuum-filtered to form 2D lamellar membranes. To prepare ion-intercalated membranes, we first fabricated neat vermiculite membranes and then filtrated ionic solutions through them. This approach preserves structural integrity during intercalation. The resulting intercalated membranes are denoted X-VM, where X represents the intercalating ions (K + , Cs + , Mg 2+ or Al 3+ ) and VM stands for vermiculite membrane (characterized in Fig. S4-S5). These X-VMs subsequently underwent deep dehydration at elevated temperature. The processed membranes were termed dp-X-VM. It is well established that most clay minerals gradually dehydrate and partially dehydroxylate over 500°C 26,27 . Consistent with this, thermogravimetric and Fourier-transform infrared spectroscopy (TG-FTIR) revealed a smooth weight loss profile below 500°C, attributable to the constant escaping of free water molecules. Above 500°C, a distinct peak emerged especially for Cs + -VM, corresponding to the loss of structurally bound water (Fig. 1 b, Fig S6-S7). Accordingly, a deep dehydration temperature of 600°C was selected to ensure complete removal of the bound water. The FTIR spectra showed the disappearance of –OH stretching vibration after deep dehydration, verifying the elimination of water molecules within dp-Cs + -VM (Fig. S8). Following deep dehydration, Cs⁺ ions became clearly visible within the dp-Cs⁺-VM interlayer structure via high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) (Fig. 1 c). In many regions, Cs + ions were sparsely distributed, creating vacant pathways conducive to ion transport (Fig. 1 d). The intercalation of Cs⁺ ions modulated the membrane’s physical properties. They partially neutralized the negative surface charges, evidenced by a less negative zeta potential (Fig. 1 e), and consequently rendering the membranes less hydrophilic (Fig. S9). This would impart the membrane with higher anti-swelling capability, as will be demonstrated later. Furthermore, deep dehydration promoted direct and tight binding between Cs⁺ ions and the vermiculite layers, potentially strengthening interlayer cohesion—another favorable trait for swelling suppression. Indeed, X-ray photoelectron spectra (XPS) of dp-Cs + -VM shows an increased binding energy of Cs4d and a new split peak of O1s, indicating stronger interaction between Cs + and vermiculite oxygen atoms 28 (Fig. 1 f, Fig. S10). Nanoindentation tests further revealed substantially higher Young’s modulus in dp-Cs⁺-VM compared to Cs⁺-VM (Fig. 1 g, Fig. S11), also suggesting the higher interlayer cohesion. This should also substantially enhance the anti-swelling capability. Enhanced anti-swelling capability To assess the anti-swelling capability, we completely immersed the membranes in pure water. The air-dried neat VM quickly disintegrated in water (Fig. 2 a). Then, the ion intercalation promoted the anti-swelling capability of X-VM, as suggested by the stable membrane form in water (Fig. S12-S14). However, they could not sustain rigorous mechanical disturbance as demonstrated by the poor stability under ultrasonication (Fig. 2 b, Fig. S15). In contrast, all dp-X-VMs maintained their membrane form either in still water or under ultrasonication. To better demonstrate the importance of the deep dehydration, we characterized the interlayer d -spacings with X-ray diffraction (XRD) (Fig. 2 c-d, Fig. S16). Without deep dehydration, the d - spacing of the K + -VM swelled from 11.27 Å to 12.70 Å, and that of the Cs + -VM swelled from 11.30 Å to 12.45 Å, representing a swell ratio of 12.7% and 10.2%, respectively. For Mg 2+ -VM and Al 3+ -VM, the swelling ratio was 23.8% and 7.4%, respectively. After deep dehydration, the swelling was suppressed for all these membranes (Fig. 2 e). The dp-K + -VM swelled by 5.2% in water, while the dp-Cs + -VM swelled even less, by merely 3.9% (Fig. 2 e). On the other hand, the swelling ratio of the dp-Mg 2+ -VM was 12.6% and that of dp-Al 3+ -VM was 15.7%. These results also demonstrate that deep dehydrated membranes intercalated with monovalent ions have higher anti-swelling capability than those with multi-valent ions (Fig. S17). This contrasts with the trend that without deep dehydration, membranes intercalated by multivalent ions were much more stable in water. We note that the sequence of the swelling ratio (Al 3+ >Mg 2+ >K + >Cs + ) coincides with that of the hydration enthalpy of these ions, suggesting that the hydration governs the swelling process. This aligns with prior studies indicating that clay swelling is initiated by the hydration of interlayer cations 29 , 30 . To better elucidate the different role of monovalent and multi-valent ions, we conducted DFT calculations. Models were constructed according to the measured d -spacings after deep dehydration. For monovalent ions, the binding energy between the intercalated ion and the VM channel exceeded the ionic hydration energy in bulk water 31 – 33 (Fig. 2 f, Table S1 ), suggesting that rehydration in dp-Cs + -VM and dp-K + -VM is energetically unfavorable, leading to their high anti-swelling performance. For multivalent ions, however, the opposite trend was found. These ions have substantially higher hydration energy, exceeding the binding energy, consequently resulting in significant swelling in water. Another key question is why deep dehydration is essential for the anti-swelling capability. As suggested by Fig. 1 b, the deep dehydration process eliminates the interlayer water molecules, thereby removing hydration forces that hinders the collapse of the interlayer channels. The reduction in interlayer free spacing should subsequently enhances interlayer cohesion, since both the attractive van der Walls interaction and the attractive cation-vermiculite interaction increases with the decrease of the interaction distance, rendering the membrane less vulnerable to swelling. Indeed, as shown in Fig. 2 g, the calculated binding energy between Cs + ion and the vermiculite channel increases steadily as interlayer free spacing decreases (Table S2). Overall, these results suggest that the interplay between the interlayer free spacing and the hydration energy of cations crucially determines the anti-swelling capability of the membrane. The deep dehydration process reduces the interlayer free spacing and thus enhances the binding energy, to an extent that surpasses the ionic hydration energy for membranes intercalated by monovalent cations. This prevents the rehydration of the interlayer cations and renders the membrane anti-swelling. Enhanced selectivity The above results also suggest that dp-Cs + -VM has the lowest swelling ratio and a narrow interlayer free spacing, making it promising for ion sieving applications. We then explored the Li + /Mg 2+ separation performance by experiments, DFT calculations, and ab initio molecular dynamics (AIMD) simulations. We chose Li + and Mg 2+ because their fast and high-selectivity separation is crucial but remains highly challenging for lithium extraction, especially from salt lake. The deep dehydration results in extremely narrow interlayer channels in water. This would significantly promote the steric hinderance against the transport of Mg 2+ ions and thus the Li + /Mg 2+ selectivity. In our AIMD simulations, we found that Mg 2+ could not enter the channels of the dp-Cs + -VM in water, while Li + spontaneously entered (Fig. 3 a-b, Fig. S18). Indeed, for Mg 2+ , the first hydration shell must shed away two water molecules to enter the VM channel, whereas the hydration shell of Li + only undergoes deformation with virtually no change in the coordination number (Fig. 3 c-d). In reality, some Mg 2+ ions would nevertheless permeate into the membrane through structural defects. Inside the VM channel, the Mg 2+ still experiences much larger energy barrier than Li + does due to the stronger binding between Mg 2+ and the VM, as suggested by our DFT calculations (Fig. 3 e, Fig. S19, Table S1 ). This guarantees that any Mg 2+ ion inside the membrane can still be hindered. The deep dehydration membranes were intercalated with many cations. This raises two questions— Will the intercalated cations leak out? How do they affect the transport of Li + and Mg 2+ ? First, as demonstrated by the AIMD simulations, the Cs + ion was trapped and could not leave the minimum-energy adsorption site as shown in the Fig. 3 a. Second, with Cs + intercalation, we found that the mean square displacement (MSD) of the Li + increased rapidly with time while that of Mg 2+ increased slowly when both ions were allowed to freely migrate in the VM channel (Fig. 3 f). In fact, when the Cs + fills the minimum-energy trap, Li + and Mg 2+ can only adsorb on higher-energy sites, leading to weakened ion-vermiculite binding and thus lower transport energy barrier (Fig. 3 g, Fig. S20-S21). Accordingly, we found that Li + ions experienced much lower energy barrier (denoted by O-Al-O/Cs + ) compared to that before Cs + intercalation (denoted by O-Al-O) (Fig. 3 h, Table S3). This could further enhance the Li + selectivity of dp-Cs + -VM. Altogether, our calculations and simulations suggest that the deep dehydrated membranes should exhibit high Li + permeability due to the reduced energy barrier after Cs + intercalation, and high Li + /Mg 2+ selectivity due to the narrow interlayer free spacing and the higher Mg 2+ -VM binding energy. Li/Mg separation for lithium extraction from salt lake Encouraged by the above results, we then evaluated the Li + /Mg 2+ separation performance of the membranes under concentration and electrical gradients, using a set of homemade dialysis or electrodialysis devices 34 (Fig. 4 a, Fig. S22). We first used a binary solution containing 0.1M LiCl and 0.1 M MgCl 2 for concentration-driven permeation test. The receiving solution was ultrapure water. In this experiment, the neat VM could not withstand prolonged tests due to its low stability in water. Following Cs + ion intercalation, the Cs + -VM exhibited moderate Li + selectivity. After deep dehydration, the Li + selectivity was substantially enhanced (Fig. 4 b, Fig. S23). For dp-Cs + -VM, we confirmed that the optimal membrane thickness was ~ 4 µm (Fig. S24). With such thickness, the Li + permeability was 0.49 mol m − 2 h − 1 and the Li + /Mg 2+ selectivity reached 92. We are aware that in practical applications, external driving force, e.g., electric voltage or pressure, is often applied to increase the permeability. Our membrane does not allow the fast transport of water under pressure (Fig. S25, Table S4), and is therefore not suitable for pressure-driven nanofiltration. However, this is just highly demanded for electrodialysis since it prevents the backflow of water and ions carried by the water flow from low concentration side to the high concentration side (Fig. S26). By tuning the externally applied voltage from 0 to 1V, we found that a moderate voltage (0.1 V) could enhance the Li + permeability and thus the Li + /Mg 2+ selectivity (Fig. 4 c, Fig. S27). Further increasing the voltage led to the decrease of Li + selectivity, presumably because such voltage is sufficient to force the divalent Mg 2+ entering the narrow channels and desorb from the vermiculite surfaces. With the optimal thickness and voltage, our membrane exhibits both high Li + permeability (0.65 mol m − 2 h − 1 ) and Mg 2+ selectivity (148), which clearly stand out among reported membranes (Fig. 4 d, Table S5). In practice, the Mg²⁺/Li⁺ molar ratio of the salt lake water vary significantly. Our membrane maintained relatively high selectivity when the Mg²⁺/Li⁺ ratio increased from 2:1 to 100:1 (Fig. S28). We also varied the pH value of the feed solutions from 4 to 10 to account for the fact that most salt-lake water are weakly acidic to weakly alkaline 35 , 36 . We found that the selectivity still kept above 80 (Fig. S29). Perhaps more interestingly, our membrane demonstrated robust acid and alkali resistance under these conditions (Fig. S30-S31), contrary to neat VM which is known to be highly vulnerable to acid and alkali. This should also be attributed to the fact that the Cs + ions are trapped on the VM surfaces, shielding them from being attacked. Remarkably, thanks to the strong anti-swelling capability, our membrane exhibited robust Li + /Mg 2+ separation performance during long time (10 days) operation (Fig. 4 e), which is crucial for industrial applications but remains rarely reported for previous high-selectivity Li + /Mg 2+ separation membranes. The high performance and high stability of the membrane allows potential practical lithium extraction applications. We used bittern from Qinghai salt lake (after solar evaporation, compositions shown in Table S6) and processed it with our membrane (Fig. 4 f, Fig. S32). By employing a one-step electrodialysis, the Mg 2+ /Li + ratio reduced from 15.1 to 0.5, and the content of the Li + increased from 6.1 wt% to 45.8 wt% (Fig. 4 g, Table S6). Then we concentrated the brines by 7.5 times, followed by precipitating the lithium with saturated Na 2 CO 3 37,38 . The final product (white powders) has a Li + purity of 99.2% (Fig. S33). XRD patterns also confirmed the obtained high purity Li 2 CO 3 (Fig. 4 h). The results demonstrate the great potential of dp-Cs + -VM membrane for highly efficient extraction of Li + from natural salt lake water. Beyond its separation performance, the membrane’s minimal water crossover under high osmotic gradients reduces operational freshwater consumption. In addition, the cost the membrane is also low thanks to the cheap raw material vermiculite and the minimal usage of CsCl (~ 1.28 g/m²; Fig. S34, Table S7). This is crucial for reducing the cost of extracting lithium resources from salt lakes, which are typically located in arid regions. Discussion The swelling of 2D layered membranes in aqueous environments has long hindered their practical application in separation processes. By exploiting the fact that the attractive van der Waals and electrostatic interaction increases with the decrease of interaction distance, we demonstrated a deep dehydration strategy to tackle this challenge, which forces the interlayer distance crossing a critical threshold that energetically disfavors rehydration and swelling. The deep dehydrated 2D layered vermiculite membranes exhibited high anti-swelling capability and narrow interlayer spacing. These enabled high Li + /Mg 2+ selectivity in long-time tests, allowing us to produce industrial-grade Li 2 CO 3 from salt lake bittern through the electrodialysis–precipitation process. Our current research focuses on the effect of deep dehydration process on membrane structure and properties, and the enhanced ion selective transport mechanism, which should shed light on design of other advanced ion sieving materials. Methods Fabrication of the vermiculite nanosheets Vermiculite nanosheets were prepared by delamination method we reported before 5 , 39 . Typically, 10 g of thermally expanded vermiculite particles (2–3 mm, purchased from Sigma-Aldrich) were dispersed in 150 mL of saturated NaCl solution and magnetically stirred for 24 h at 110°C for sufficient cation exchange. After repeated centrifugation washing with ethanol and water, the sediment was dispersed in 150 mL of 2 M LiCl solution and magnetically stirred for 24 h at 110°C. The Li + -exchanged vermiculite was then obtained after the same washing procedure as mentioned above. Then, the vermiculite slurry was mixed with 150 mL of 30% H 2 O 2 solution and magnetically stirred at 110°C for 24 h. The mixture was then ultrasonically treated 1 h for further exfoliation, followed by centrifugation at 6000 rpm for 10 min, to obtain the colloidal dispersion of monolayered nanosheets. The total mass of the resulting nanosheets is about 700 to 1000 mg. Fabrication of the X-VM and dp-X-VM membranes The lamellar VM membranes were assembled by the nanosheets via vacuum-filtrating. Typically, a certain amount of VM nanosheets were dispersed in at least 100 mL ultrapure water, then filtrated onto a PP substrate until the water was drained out. Afterwards, the VM membrane with PP substrate was heated at 60°C for 8 h, then free-standing VM membrane could be peeled from the substrate. For ion intercalation, 50 mL of 10 − 3 M CsCl (or KCl, MgCl 2 , and AlCl 3 ) solution was added to the surface of a freshly assembled VM membrane. The filtration process was then continued until there is no visible water staining on the membrane surface. Given the low consumption of intercalation cations (Fig. S34, Table S7), the chloride salt solution could be collected to reuse. The deep dehydration process of VM membranes was enabled by programmed calcination in tube furnace (TL-1200, Nanjing Boyuntong Instruments, China) under inert atmosphere. The heat up and cool down rates are set to 10°C/min. Holding for 1 h after heating up to the specified temperature to ensure uniform dehydration at that temperature. The argon gas flow rate through the furnace was ca. 300 cm 3 min − 1 . Characterization Surface and cross-sectional morphology characterizations were carried out with a field-emission scanning electron microscopy (Hitachi SU-8010, Japan) integrated with energy-dispersive X-ray spectroscopy (EDS) and atomic force microscopy (AFM, Agilent 5400, USA). The lattice structure and atomic arrangement in interlayer spacings were obtained by scanning transmission election microscope (STEM, Thermo Scientific Themis Z, USA) with high-angle annular dark-field (HAADF). The surface charge property of the nanosheets and membrane samples was analyzed by zeta potential analyzer (Malvern Zetasizer Nano ZS90, UK) and electrokinetic analyzer (Anton Paar SurPASS3, Austria) respectively. The surface chemical properties were characterized by Fourier transform infrared spectrometry (FT-IR, Nicolet 6700, USA) and X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA). The dehydration process of the materials was characterized by thermogravimetric analysis and Fourier-transform infrared spectroscopy coupled technology (TGA-FTIR, Netzsch STA-2500, Germany and Thermo Fisher IS-50, USA). XRD patterns were measured using X-ray diffractometer (Rigaku SmartLab 9KW, Japan). The water contact angle was measured using a contact angle goniometer (LAUDA Scientific LSA 100, Germany). The mechanical properties are tested with nano indenter (Bruker Hysitron TI 980, Germany) for three different area of one sample. Ion selectivity and lithium extraction The diffusion dialysis experiments were realized by the self-made permeation apparatus with two polytetrafluoroethylene (PTFE) reservoirs. The tested membrane (effective area 7.07 mm 2 ) was mounted in the center to separate two sides. During the experiments, one reservoir was filled with 10 ml feed solution containing LiCl and MgCl 2 (0.1 M unless otherwise specified) while 10 ml ultrapure water was added in the permeation reservoir. For pH-stability permeation tests, trace amounts of NaOH or HCl solution were used to adjust the pH of the feed solution. For the external electric field driven permeation process, inert graphite plate electrodes (30×10×1 mm) were introduced in the reservoir, and a DC power supply was used to control the bias voltage between the two sides. The electrode in the high concentration side was connected to the positive pole of the power supply. The salt lake bittern was obtained from Qinghai Lake, China (Table S6). The continuous electrodialysis was performed using homemade setups (Fig. S32). The lithium-rich solution after electrodialysis was further treated to obtain Li 2 CO 3 product 37 . In general, a saturated Na 2 CO 3 solution was added to the collected lithium-rich solution at 80°C 38 . The obtained white sediments can be further refined by dissolving in ultrapure water under 20°C and then heated to 90°C to filter the precipitate while hot. The generated Li 2 CO 3 were collected and washed by centrifuge and then dried in an air-forced oven at 70°C for at least 12 h. The sediments were dissolved in dilute hydrochloric acid solutions to determine the lithium purity. The concentration of cations in the permeated solution and the salt lake water were measured by inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7850, USA) for ppb level of concentration and inductively coupled plasma atomic emission spectroscopy (ICP-OES, Agilent 5800, USA) for ppm level of concentration. The anion permeation concentration was measured by ion chromatograph (Metrohm 863 Basic IC plus, Switzerland). At least three parallel samples were tested to evaluate systematic errors. The Li + /Mg 2+ selectivity in the solution was calculated with molar concentrations by the following Eq. ( 1 ), $$\:{S}_{{\text{L}\text{i}}^{+}/{\text{M}\text{g}}^{2+}}\:=\:\frac{{\left({C}_{{\text{L}\text{i}}^{+}}/{C}_{{\text{M}\text{g}}^{2+}}\right)}_{dialysate}}{{\left({{C}_{{\text{L}\text{i}}^{+}}/C}_{{\text{M}\text{g}}^{2+}}\right)}_{feed\:solution}}$$ 1 Density functional theory (DFT) calculation The DFT calculations of non-periodic cluster models were performed on ORCA 6.0.0 program packages 42 . The optimized structures were conducted at B3LYP level of theory with def2-TZVP basis set 43 . All calculations have contained the Grimme's D3 dispersion correction (BJ-damping) and taken the solvent effect into account 44 . Periodic DFT calculations were conducted on CP2K 2024.1 package 45 . The VM single unit cell (lattice parameters a = 5.349 Å, b = 9.255 Å, c = 28.89 Å and α = γ = 90.00°, β = 97.12°) was expanded by replication into a 2×2×1 supercell with interlayer-filled magnesium ions removed. The number of water molecules filled at different interlayer d -spacings was referenced from literature reports 30 . The vacuum layer thickness is large enough to avoid any possible interaction between the repeated slabs. Perdew-Burke-Ernzerhof (PBE) with Grimme’s D3 dispersion corrections were employed for the exchange and correlation functional 44 , 46 . The core electrons were treated using the Goedecker-Tetter-Hutter (GTH) pseudo-potentials, and the valence electrons were expanded using the Gaussian and Plane-Wave (GPW) combined basis sets (DZVP-MOLOPT for structural optimization and vibrational analysis, TZVP for energy calculation with 400 Ry cut-off energy) 47 , 48 . The Monkhorst-Pack method was used to generate the Brillouin zone integration from a Gamma centered 4×2×1 k-point mesh. All wave function analyses (including non-covalent interaction using independent gradient model based on Hirshfeld partition 49 , Hirshfeld-I atomic charges 50 ) and visualizations were done with the help of Multiwfn (3.8dev) 51 and VMD (1.9.3) software 52 . The energy barrier for a single ion transporting through the different sites of the channel using the climbing image nudged elastic band (CI-NEB) method 53 . All binding energy calculations take the basis set superposition error (BSSE) into consideration 54 , thus the interaction of a given ion (M n+ ) with VM nanosheet is calculated as Eq. ( 2 ), $$\:{E}_{Binding}={E}_{{M}^{n+}@VM}-{E}_{{M}^{n+}}-{E}_{VM}+{E}_{BSSE}$$ 2 The Shermo program was used to calculate thermodynamic data, the system temperature and pressure were set to 298.15 K and 1 atm respectively 55 . Ab-initio molecular dynamics (AIMD) simulation AIMD calculations were performed to study the structure and mobility of hydrated ions in VM nanochannel using the same theoretical level of DFT calculations with CP2K. The temperature was fixed at 298.15 K using the Nosé-Hoover thermostat with a damping constant of 100 fs. Periodic boundary conditions were applied in all directions. The mean square displacement (MSD) was calculated by Eq. ( 3 ), $$\:MSD=\frac{1}{N}\sum\:_{i=1}^{N}{\left[{r}_{i}\left(t\right)-{r}_{i}\left(0\right)\right]}^{2}$$ 3 where \(\:N\) is the total number of observed species in the unit cell, and \(\:{r}_{i}\left(t\right)\) is the position of the \(\:i\) -th species at time \(\:t\) . Then the self-diffusion coefficient ( D ) of the cation was fitted from the slope of MSD curves as Eq. ( 4 ), $$\:D=\frac{1}{6}\underset{t\to\:\infty\:}{\text{lim}}\frac{MSD\left(t\right)}{t}$$ 4 For the non-equilibrium simulations of the entrance of ions into the nanochannels, a small velocity (0.002 Å fs − 1 ) was used on cations to perform constrained AIMD, which could mimic the migration of ions under a concentration gradient. The nanosheet was fixed throughout the simulation in order to collect the trajectories of specific ions. Declarations Data Availability The data supporting the findings of this study are available within the paper and its Supplementary Information. Source data are provided with this paper. Acknowledgments This work was supported by the National Key R&D Program of China (NO. 2022YFB3805900), the National Natural Science Foundation of China (22272194), Shandong Provincial Natural Science Foundation (ZR2021YQ12, ZR2024QB294), QIBEBT/SEI/QNESL (S202303), Key R&D Program of Shandong Province (2022CXGC010302) and the China Postdoctoral Science Foundation (No. 2024M753353). The authors acknowledge the use of DeepSeek v3.1 to correct language errors. Author Contributions Z.M. conducted the experiments and simulations with help from J.F., C.Z. and H.G., Z.M. wrote and revised the draft. J.G. conceptualized and supervised the project and revised the draft. Competing Interest s The authors declare no competing interests. Additional information Supplementary information is available online. Correspondence and requests for materials should be addressed to J. G. References Coleman, J. N. et al. Two-Dimensional Nanosheets Produced by Liquid Exfoliation of Layered Materials. Science 331 , 568–571 (2011). Gao, J., Feng, Y., Guo, W. & Jiang, L. Nanofluidics in two-dimensional layered materials: inspirations from nature. Chem. Soc. Rev. 46 , 5400–5424 (2017). Cai, X., Luo, Y., Liu, B. & Cheng, H.-M. Preparation of 2D material dispersions and their applications. Chem. Soc. Rev. 47 , 6224–6266 (2018). Kang, Y., Xia, Y., Wang, H. & Zhang, X. 2D Laminar Membranes for Selective Water and Ion Transport. Adv. Funct. Mater. 29 , 1902014 (2019). Huang, T. et al. Two-Dimensional Sodium Channels with High Selectivity and Conductivity for Osmotic Power Generation from Wastewater. ACS Nano 17 , 17245–17253 (2023). Yang, R. et al. Synthesis of atomically thin sheets by the intercalation-based exfoliation of layered materials. Nat. Synth. 2 , 101–118 (2023). Yeh, C.-N., Raidongia, K., Shao, J., Yang, Q.-H. & Huang, J. On the origin of the stability of graphene oxide membranes in water. Nat. Chem. 7 , 166–170 (2015). Zheng, S., Tu, Q., Urban, J. J., Li, S. & Mi, B. Swelling of Graphene Oxide Membranes in Aqueous Solution: Characterization of Interlayer Spacing and Insight into Water Transport Mechanisms. ACS Nano 11 , 6440–6450 (2017). Liu, Y. et al. Pillared Laminar Vermiculite Membranes with Tunable Monovalent and Multivalent Ion Selectivity. Adv. Mater. 2417994 (2025) doi:10.1002/adma.202417994. Lu, Z. et al. Self-Crosslinked MXene (Ti 3 C 2 T x ) Membranes with Good Antiswelling Property for Monovalent Metal Ion Exclusion. ACS Nano 13 , 10535–10544 (2019). Ding, L. et al. Effective ion sieving with Ti3C2Tx MXene membranes for production of drinking water from seawater. Nat. Sustain. 3 , 296–302 (2020). Helal, M. I. et al. Understanding the swelling behavior of Ti 3 C 2 T x MXene membranes in aqueous media. J. Mater. Chem. A 12 , 30729–30742 (2024). Chen, L. et al. Ion sieving in graphene oxide membranes via cationic control of interlayer spacing. Nature 550 , 380–383 (2017). Wang, W. et al. 2D Vermiculite Nanolaminated Membranes for Efficient Organic Solvent Nanofiltration. Adv. Funct. Mater. 2410635 (2024) doi:10.1002/adfm.202410635. Hribar, B., Southall, N. T., Vlachy, V. & Dill, K. A. How ions affect the structure of water. J. Am. Chem. Soc. 124 , 12302–12311 (2002). Leite, F. L., Bueno, C. C., Da Róz, A. L., Ziemath, E. C. & Oliveira, O. N. Theoretical models for surface forces and adhesion and their measurement using atomic force microscopy. Int. J. Mol. Sci. 13 , 12773–12856 (2012). Duong, D. L., Yun, S. J. & Lee, Y. H. van der waals layered materials: Opportunities and challenges. ACS Nano 11 , 11803–11830 (2017). Vasu, K. S. et al. Van der waals pressure and its effect on trapped interlayer molecules. Nat. Commun. 7 , (2016). Motokawa, R. et al. Collective Structural Changes in Vermiculite Clay Suspensions Induced by Cesium Ions. Sci. Rep. 4 , 6585 (2014). Yin, X., Takahashi, H., Inaba, Y. & Takeshita, K. Desorption of Cesium Ions from Vermiculite with Sea Water by Hydrothermal Process. Chem. Lett. 45 , 256–258 (2016). Razmjou, A., Asadnia, M., Hosseini, E., Habibnejad Korayem, A. & Chen, V. Design principles of ion selective nanostructured membranes for the extraction of lithium ions. Nat. Commun. 10 , 5793 (2019). Hou, L. et al. Understanding the Ion Transport Behavior across Nanofluidic Membranes in Response to the Charge Variations. Adv. Funct. Mater. 31 , 2009970 (2021). Zhang, Y. et al. Congener-welded crystalline carbon nitride membrane for robust and highly selective Li/Mg separation. Sci. Adv. 10 , eadm9620 (2024). Li, Z. et al. Lithium extraction from brine through a decoupled and membrane-free electrochemical cell design. Science 385 , 1438–1444 (2024). Yang, S., Wang, Y., Pan, H., He, P. & Zhou, H. Lithium extraction from low-quality brines. Nature 636 , 309–321 (2024). Fernandez, R., Martirena, F. & Scrivener, K. L. The origin of the pozzolanic activity of calcined clay minerals: A comparison between kaolinite, illite and montmorillonite. Cem. Concr. Res. 41 , 113–122 (2011). Xie, W. et al. Evolution of the crystallographic structure and physicochemical aspects of rectorite upon calcination. Appl. Clay Sci. 185 , 105374 (2020). Dupin, J.-C., Gonbeau, D., Vinatier, P. & Levasseur, A. Systematic XPS studies of metal oxides, hydroxides and peroxides. Phys. Chem. Chem. Phys. 2 , 1319–1324 (2000). Hensen, E. J. M. & Smit, B. Why clays swell. J. Phys. Chem. B 106 , 12664–12667 (2002). Teich-McGoldrick, S. L., Greathouse, J. A., Jové-Colón, C. F. & Cygan, R. T. Swelling properties of montmorillonite and beidellite clay minerals from molecular simulation: Comparison of temperature, interlayer cation, and charge location effects. J. Phys. Chem. C 119 , 20880–20891 (2015). Randles, J. E. B. The real hydration energies of ions. Trans. Faraday Soc. 52 , 1573–1581 (1956). Kollman, P. A. & Kuntz, I. D. Studies of Cation hydration. J. Am. Chem. Soc. 94 , 9236–9237 (1972). Kepp, K. P. Free energies of hydration for metal ions from heats of vaporization. J. Phys. Chem. A 123 , 6536–6546 (2019). Bao, S. et al. Randomly oriented covalent organic framework membrane for selective li+ sieving from other ions. Nat. Commun. 16 , 3896 (2025). Zachara, J. M. et al. Geo- and biogeochemical processes in a heliothermal hypersaline lake. Geochim. Cosmochim. Acta 181 , 144–163 (2016). Ghori, N.-U.-H., Wise, M. J. & Whiteley, A. S. Temporal Microbial Community Dynamics Within a Unique Acid Saline Lake. Front. Microbiol. 12 , 649594 (2021). Linneen, N., Bhave, R. & Woerner, D. Purification of industrial grade lithium chloride for the recovery of high purity battery grade lithium carbonate. Sep. Purif. Technol. 214 , 168–173 (2019). Zhang, S. et al. Solar-driven membrane separation for direct lithium extraction from artificial salt-lake brine. Nat. Commun. 15 , 238 (2024). Shao, J.-J., Raidongia, K., Koltonow, A. R. & Huang, J. Self-assembled two-dimensional nanofluidic proton channels with high thermal stability. Nat. Commun. 6 , 7602 (2015). Rettig, S. L., Jones, B. F. & Risacher, F. Geochemical evolution of brines in the Salar of Uyuni, Bolivia. Chem. Geol. 30 , 57–79 (1980). An, J. W. et al. Recovery of lithium from Uyuni salar brine. Hydrometallurgy 117–118 , 64–70 (2012). Neese, F. Software update: The ORCA program system—Version 5.0. WIREs Comput. Mol. Sci. 12 , e1606 (2022). Weigend, F. & Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. Phys. Chem. Chem. Phys. 7 , 3297 (2005). Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132 , 154104 (2010). Kühne, T. D. et al. CP2K: An electronic structure and molecular dynamics software package - Quickstep: Efficient and accurate electronic structure calculations. J. Chem. Phys. 152 , 194103 (2020). Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 77 , 3865–3868 (1996). Goedecker, S., Teter, M. & Hutter, J. Separable dual-space Gaussian pseudopotentials. Phys. Rev. B 54 , 1703–1710 (1996). Krack, M. Pseudopotentials for H to Kr optimized for gradient-corrected exchange-correlation functionals. Theor. Chem. Acc. 114 , 145–152 (2005). Lu, T. & Chen, Q. Independent gradient model based on Hirshfeld partition: A new method for visual study of interactions in chemical systems. J. Comput. Chem. 43 , 539–555 (2022). Bultinck, P., Van Alsenoy, C., Ayers, P. W. & Carbó-Dorca, R. Critical analysis and extension of the Hirshfeld atoms in molecules. J. Chem. Phys. 126 , 144111 (2007). Lu, T. & Chen, F. Multiwfn: A multifunctional wavefunction analyzer. J. Comput. Chem. 33 , 580–592 (2012). Humphrey, W., Dalke, A. & Schulten, K. VMD: Visual molecular dynamics. J. Mol. Graph. 14 , 33–38 (1996). Henkelman, G., Uberuaga, B. P. & Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113 , 9901–9904 (2000). Kestner, N. R. & Combariza, J. E. Basis Set Superposition Errors: Theory and Practice. in Reviews in Computational Chemistry 99–132 (1999). doi:10.1002/9780470125908.ch2. Lu, T. & Chen, Q. Shermo: A general code for calculating molecular thermochemistry properties. Comput. Theor. Chem. 1200 , 113249 (2021). Additional Declarations There is NO Competing Interest. Supplementary Files NCSI.docx Supplementary figures and tables Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7726847","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":527288235,"identity":"9b10d4e8-cd09-4f5a-8a62-1bfae23110aa","order_by":0,"name":"Jun Gao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYBACxgYILcfAzMPATJIWY4iWBBJsS2xgIFYL84zcA4xfamzS+47zHnxc+INBnr+B+dkDvA6bkZfALHMsLXfmYb5k4xkJDIYzDrCZG+DXkmPALMF2OHfDYR4zaZ4EBsYNDDxsEoS1/DucbnCYx/w3UIs9UVoYP7YdTgBqMWMGakkkrKXnjcFhxr40Q5BfpGekSSTPOMxmhleLYXuO4cMf32zk+c6fPfi5wMbGtr+9+Rl+LQ0MDId5QKwDYD5QMaE0IA9y3A+EllEwCkbBKBgFmAAANYtB320eA5wAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-2106-3134","institution":"Qingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences","correspondingAuthor":true,"prefix":"","firstName":"Jun","middleName":"","lastName":"Gao","suffix":""},{"id":527288236,"identity":"e9432e85-f39f-4b3d-8a0f-31b87e3c7d9a","order_by":1,"name":"Zhaoyu Ma","email":"","orcid":"https://orcid.org/0000-0001-6747-5359","institution":"Qingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Zhaoyu","middleName":"","lastName":"Ma","suffix":""},{"id":527288237,"identity":"558b4cf6-38f4-406e-a2b6-d2fde507a5aa","order_by":2,"name":"Jilong Fan","email":"","orcid":"","institution":"Qingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Jilong","middleName":"","lastName":"Fan","suffix":""},{"id":527288238,"identity":"bb244a89-45a4-4ad1-a057-c8d493f4c88e","order_by":3,"name":"Chenguang Zhu","email":"","orcid":"https://orcid.org/0000-0002-8189-5592","institution":"Qingdao Institute of Bioenergy \u0026 Bioprocess Technology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Chenguang","middleName":"","lastName":"Zhu","suffix":""},{"id":527288239,"identity":"71942234-6f6c-4754-b3cc-94bc2620517c","order_by":4,"name":"Hongfei Gao","email":"","orcid":"","institution":"Qingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Hongfei","middleName":"","lastName":"Gao","suffix":""}],"badges":[],"createdAt":"2025-09-27 07:50:11","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7726847/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7726847/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93400877,"identity":"3a2e745c-3acf-4645-a513-2628511545cf","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4178742,"visible":true,"origin":"","legend":"","description":"","filename":"NCmanuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/4c365147ffc5744cf376b67d.docx"},{"id":93401574,"identity":"01694895-5be9-44db-97f9-91c07ebfea94","added_by":"auto","created_at":"2025-10-13 12:40:25","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6599,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS2577525.json","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/4ac2e810288599e5d50697b2.json"},{"id":93400890,"identity":"6bdb32ef-86f8-4d44-b431-6c740e2eccc3","added_by":"auto","created_at":"2025-10-13 12:32:26","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":27523230,"visible":true,"origin":"","legend":"","description":"","filename":"NCSI.docx","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/85545ea661adaf9896c5ddc5.docx"},{"id":93400879,"identity":"f352ca89-0bc0-4350-9614-2f7a2cfd6144","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121585,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25775250enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/b531d5f97524a8d6f7a06cb7.xml"},{"id":93400878,"identity":"3504b638-9931-497f-9bc0-22ce8eeee9ed","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1070049,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/e47e1f94bfdb2f3abb92b790.jpeg"},{"id":93400882,"identity":"6eb90055-d208-4569-ae6e-791ebac49324","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1033963,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/7f796f71af7b7f472be639e6.jpeg"},{"id":93400880,"identity":"084b2f2f-6669-4008-a0f0-0d99b6e03c1c","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1036937,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/0eebd929564f10ad822b2288.jpeg"},{"id":93401576,"identity":"448b3103-83dc-4d01-8d84-4f15a0a23ac1","added_by":"auto","created_at":"2025-10-13 12:40:25","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":957840,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/08db9307a593c017be6d9196.jpeg"},{"id":93400885,"identity":"170a5cf3-0939-4960-b1db-14b79e4ac585","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":405433,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/a04b26edc95593c04e93a868.png"},{"id":93400886,"identity":"fce0da6c-8b66-4cad-84b1-f325fdadb267","added_by":"auto","created_at":"2025-10-13 12:32:26","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":376468,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/a9240eb43b8d39d0c1ca35c2.png"},{"id":93401578,"identity":"2e351f23-0a95-44be-a4dd-900d6322dde0","added_by":"auto","created_at":"2025-10-13 12:40:26","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":347322,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/4ab337961964de27ac989bf5.png"},{"id":93400883,"identity":"2e4695bc-b211-4f0e-865e-e59d695a3439","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":348759,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/4d00c3847cde7b7e32828902.png"},{"id":93400887,"identity":"90fff674-5087-4a5d-9e29-c5d8f8f81ce0","added_by":"auto","created_at":"2025-10-13 12:32:26","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117319,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25775250structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/397b1374a82a5f5dce8661d5.xml"},{"id":93401577,"identity":"f1c88c04-e647-4c20-bcf0-f1b75075a6af","added_by":"auto","created_at":"2025-10-13 12:40:25","extension":"html","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":126954,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/158d5fe44a0a0d0234b39239.html"},{"id":93400872,"identity":"0c99d33a-9dab-413a-a91f-06e231989f43","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1070049,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFabrication and characterization of the X-VM and dp-X-VM.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e) Schematic illustration of the construction of the ion-intercalated X-VM and deep-dehydrated dp-X-VM membranes. (\u003cstrong\u003eb\u003c/strong\u003e) Thermogravimetric and Fourier-transform infrared spectroscopy (TG-FTIR) coupled analysis of the Cs\u003csup\u003e+\u003c/sup\u003e-VM membrane. The heating rate is 10 °C/min. The arrows indicate a massive release peak of bound water beyond 500 °C (as escaping gaseous water with characteristic absorption peak at 1300–1500 cm\u003csup\u003e-1\u003c/sup\u003e and 3500–3950 cm\u003csup\u003e-1\u003c/sup\u003e). (\u003cstrong\u003ec\u003c/strong\u003e) High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images of the cross-section of dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM membrane. (\u003cstrong\u003ed\u003c/strong\u003e) HAADF-STEM top view of the dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM surface show that the distribution region of Cs\u003csup\u003e+\u003c/sup\u003e ions is discrete, providing pathways to transport ions (dotted lines). (\u003cstrong\u003ee\u003c/strong\u003e) Zeta potential of the VM, dp-VM, Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM at pH 6.5. (\u003cstrong\u003ef\u003c/strong\u003e) XPS fine scan spectra of Al2p and Cs4d of neat VM, Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM. (\u003cstrong\u003eg\u003c/strong\u003e) Depth-load curves of Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM during nanoindentation tests. Inset is the resulted Young's modulus (Er). Error bars represent standard deviation of the data from at least three individual experiments.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/9caae06de41863bf131ddca5.jpeg"},{"id":93400875,"identity":"624fe348-972d-4dcb-a0f5-9d17b2ffc1c2","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1033963,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDeep dehydrated ion-intercalated VM with enhanced anti-swelling stability.\u003c/strong\u003e (\u003cstrong\u003ea\u003c/strong\u003e) Comparison of the neat VM, Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM soaking in ultrapure water. (\u003cstrong\u003eb\u003c/strong\u003e) High-power ultrasonic impact test of Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM. 5 min treatment under 800 W. (\u003cstrong\u003ec\u003c/strong\u003e) XRD patterns of the VM, Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM. (\u003cstrong\u003ed\u003c/strong\u003e) The \u003cem\u003ed\u003c/em\u003e-spacings of dp-X-VMs dehydrated at 600 °C in dry or wet state. The interlayer free spacing is defined as the \u003cem\u003ed\u003c/em\u003e-spacing in wet state minus the theoretical thickness of vermiculite monolayer. (\u003cstrong\u003ee\u003c/strong\u003e) The swelling ratio of the neat VM and X-VMs before and after deep dehydration. (\u003cstrong\u003ef\u003c/strong\u003e) Comparison of the hydration free energy of the intercalated ions and their binding energy between vermiculite nanosheet. (\u003cstrong\u003eg\u003c/strong\u003e) The variation of the binding energy of Cs\u003csup\u003e+\u003c/sup\u003e and vermiculite double layer with a reduced interlayer free spacing during dehydration process.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/0e110ba69d1ea2ea2bf8a8a3.jpeg"},{"id":93400874,"identity":"fc043fa9-f31c-4c42-bfe8-4bec5b6847d8","added_by":"auto","created_at":"2025-10-13 12:32:25","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1036937,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe mechanism of enhanced ion selectivity.\u003c/strong\u003e (\u003cstrong\u003ea-b\u003c/strong\u003e) Trajectories of Li\u003csup\u003e+\u003c/sup\u003e (blue line) and Mg\u003csup\u003e2+\u003c/sup\u003e (orange line) ions at the entrance of the vermiculite bilayer for 20 ps of AIMD simulation. The trajectory of Cs\u003csup\u003e+\u003c/sup\u003e ion (grey line) shows its stable adsorption in the channel. (\u003cstrong\u003ec-d\u003c/strong\u003e) Water radial distribution function (RDF) g(r) curves (solid line) and corresponding water coordination number (dashed line) of Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e in bulk and in VM channel. (\u003cstrong\u003ee\u003c/strong\u003e) Binding energy between hydrated Li\u003csup\u003e+\u003c/sup\u003e, Mg\u003csup\u003e2+\u003c/sup\u003e and vermiculite nanosheet. Insets are the configurations of adsorbed hydrated ions. (\u003cstrong\u003ef\u003c/strong\u003e) The mean square displacement (MSD) curves of ions within VM channel (inset is the screenshot of the AIMD simulation process). The interlayer free spacing was fixed at \u003cem\u003eca.\u003c/em\u003e 3 Å. (\u003cstrong\u003eg\u003c/strong\u003e) Three ion transport pathways with different potential wells for Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e. (\u003cstrong\u003eh\u003c/strong\u003e) Calculated diffusion energy barriers of Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+ \u003c/sup\u003eions. The presence of Cs\u003csup\u003e+\u003c/sup\u003e ion can reduce the transport energy barrier of Li\u003csup\u003e+\u003c/sup\u003e ion.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/70636f7365e763d76321dc67.jpeg"},{"id":93401575,"identity":"661f22b8-0677-4eaa-a59b-c956ecb20092","added_by":"auto","created_at":"2025-10-13 12:40:25","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":957840,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLi\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e/Mg\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e separation for lithium extraction from salt lake. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Schematic illustration of the permeation test and ion selective transport. (\u003cstrong\u003eb\u003c/strong\u003e) Concentration-gradient-driven ion sieving performances. (\u003cstrong\u003ec\u003c/strong\u003e) Effect of the external voltage during permeation. (\u003cstrong\u003ed\u003c/strong\u003e) Comparison of Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity of our work with recent reports. (\u003cstrong\u003ee\u003c/strong\u003e) Stability of the dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM membrane for a long-term permeation test. (\u003cstrong\u003ef\u003c/strong\u003e) The two-step process to obtain lithium salt product from a salt lake bittern. The lithium content was determined as a percentage of the total five main cations (Li\u003csup\u003e+\u003c/sup\u003e, Na\u003csup\u003e+\u003c/sup\u003e, K\u003csup\u003e+\u003c/sup\u003e, Mg\u003csup\u003e2+\u003c/sup\u003e, Ca\u003csup\u003e2+\u003c/sup\u003e) of the bittern. (\u003cstrong\u003eg\u003c/strong\u003e) Optical photos of the bittern, the dialysate after electrodialysis, the concentrated dialysate, and the white sediments generating within concentrated dialysate after adding saturated Na\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e solution (Detailed extraction steps in Methods and detailed compositions in Supplementary information) (\u003cstrong\u003eh\u003c/strong\u003e) Powder XRD pattern of the refined sediment (inset photo) indicates a good match with the standard pattern of Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e. Error bars in all cases\u003cstrong\u003e \u003c/strong\u003erepresent standard deviation of the data from at least three individual experiments.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/61ebd6a881feb0069d164d00.jpeg"},{"id":102295421,"identity":"d7e3a066-2eca-4d38-b359-997ee7d24802","added_by":"auto","created_at":"2026-02-10 10:11:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":5086388,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/606ed124-aa1f-4655-9529-2ffc2fe4b975.pdf"},{"id":93400889,"identity":"407fbdc6-6dd7-42a0-bd23-6aadd535e042","added_by":"auto","created_at":"2025-10-13 12:32:26","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":27523230,"visible":true,"origin":"","legend":"Supplementary figures and tables","description":"","filename":"NCSI.docx","url":"https://assets-eu.researchsquare.com/files/rs-7726847/v1/e9324f7ad1c25f5bb09750d6.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Deep dehydrated layered membrane for selective and fast ion separation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eTwo-dimensional (2D) layered membranes, obtained by assembling the 2D nanosheets exfoliated from their bulk form, are regarded highly promising materials for molecular separation due to their facile fabrication method, high scalability, as well as the high structural and chemical tunability\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. However, their practical application is severely hindered by their compromised selectivity upon swelling in solvents, particularly in water. For example, neat graphene oxide and vermiculite layered membranes easily disintegrate in water\u003csup\u003e\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, and neat MXene membranes swell by over 20% in water\u003csup\u003e\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eTo address this problem, we should first revisit the mechanism of swelling. For layered materials, it is known that the attractive Coulomb interaction mediated by the intercalated counter-ions and the Van der Waals interaction between nanosheet layers prevent swelling while the hydration of the intercalated counter-ions initiates swelling\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Consequently, the most common practice to mitigate swelling is to cross-link the membranes with counter-ions, especially with multivalent ions which has stronger Coulomb interaction\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Yet, such method experienced mixed success because multivalent ions also have larger hydration enthalpy and still cause a large swelling ratio. Monovalent ions, on the other hand, exhibit weak Coulomb interaction and can be easily exchanged out during the separation process, losing the cross-linking capability.\u003c/p\u003e\u003cp\u003eWe recall that both Coulomb interaction and Van der Waals interaction steeply increase with the decrease of the interaction distance, whereas the hydration interaction does not\u003csup\u003e\u003cspan additionalcitationids=\"CR16\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. This suggests that if we squeeze the interlayer spacing to an extreme extent, the attractive interaction could overtop the hydration interaction and thus prevent the swelling. It is of course difficult to mechanically squeeze the spacing, since in the atomic scale, the interlayer pressure is up to the GPa scale\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. We here present a simple but effective method to circumvent this thorny issue. We dehydrate the layered membrane at high temperature (\u0026gt;\u0026thinsp;500\u0026deg;C), a process we term deep dehydration, to remove the free and hydration water molecules in the interlayer spacings. In this way, the hydration interaction is eliminated, and the attractive interactions naturally compress the membrane, forcing the interlayer spacing size crossing the anti-swelling threshold value.\u003c/p\u003e\u003cp\u003eWe demonstrated the method with alkaline-ion-intercalated vermiculite membranes (VM). The deep dehydration resulted in an extremely narrow \u003cem\u003ed\u003c/em\u003e-spacing (10.5\u0026thinsp;~\u0026thinsp;11.5 \u0026Aring;, representing a free spacing of 1.5\u0026thinsp;~\u0026thinsp;2.5 \u0026Aring; after subtracting the theoretical monolayer thickness of 0.9 nm). Density functional theory (DFT) calculations reveal that during the compression of the interlayer channel, the intercalated ions were tightly trapped into the minimum-energy positions, hindering their rehydration. More importantly, the extremely narrow interlayer channel greatly amplifies the binding between the intercalated alkaline ions and the vermiculite, rendering the rehydration energetically unfavorable.\u003c/p\u003e\u003cp\u003eAs a result, the membrane only swells by 0.97 \u0026Aring; (intercalated by K\u003csup\u003e+\u003c/sup\u003e) or 0.45 \u0026Aring; (intercalated by Cs\u003csup\u003e+\u0026thinsp;19,20\u003c/sup\u003e, which has lower hydration energy than K\u003csup\u003e+\u003c/sup\u003e) even after immersed in pure water for 30 days, whereas the membranes intercalated by the same alkaline ions without deep dehydration quickly swell in water in a few hours. Thanks to the stable and extremely narrow interlayer free spacing, the deep dehydrated Cs\u003csup\u003e+\u003c/sup\u003e-intercalated membrane is able to separate Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e with high selectivity (148), which is crucial but challenging for salt lake lithium extraction\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23 CR24\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Interestingly, the permeability of the Li\u003csup\u003e+\u003c/sup\u003e also reached a high value of 0.65 mol m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e despite the confinement. We attribute to the fact that the intercalation ions filled the minimum-energy traps, thereby preventing the Li\u003csup\u003e+\u003c/sup\u003e ions from being trapped and reducing their transport resistance.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eFabrication and characterization of the deep-dehydrated membranes\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea illustrates the fabrication process based on ion intercalation and deep dehydration. Vermiculite nanosheets (characterized in Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-S3) exfoliated from the bulk crystals, were dispersed in water and vacuum-filtered to form 2D lamellar membranes. To prepare ion-intercalated membranes, we first fabricated neat vermiculite membranes and then filtrated ionic solutions through them. This approach preserves structural integrity during intercalation. The resulting intercalated membranes are denoted X-VM, where X represents the intercalating ions (K\u003csup\u003e+\u003c/sup\u003e, Cs\u003csup\u003e+\u003c/sup\u003e, Mg\u003csup\u003e2+\u003c/sup\u003e or Al\u003csup\u003e3+\u003c/sup\u003e) and VM stands for vermiculite membrane (characterized in Fig. S4-S5). These X-VMs subsequently underwent deep dehydration at elevated temperature. The processed membranes were termed dp-X-VM.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIt is well established that most clay minerals gradually dehydrate and partially dehydroxylate over 500\u0026deg;C\u003csup\u003e26,27\u003c/sup\u003e. Consistent with this, thermogravimetric and Fourier-transform infrared spectroscopy (TG-FTIR) revealed a smooth weight loss profile below 500\u0026deg;C, attributable to the constant escaping of free water molecules. Above 500\u0026deg;C, a distinct peak emerged especially for Cs\u003csup\u003e+\u003c/sup\u003e-VM, corresponding to the loss of structurally bound water (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, Fig S6-S7). Accordingly, a deep dehydration temperature of 600\u0026deg;C was selected to ensure complete removal of the bound water. The FTIR spectra showed the disappearance of \u0026ndash;OH stretching vibration after deep dehydration, verifying the elimination of water molecules within dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM (Fig. S8).\u003c/p\u003e\u003cp\u003eFollowing deep dehydration, Cs⁺ ions became clearly visible within the dp-Cs⁺-VM interlayer structure via high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). In many regions, Cs\u003csup\u003e+\u003c/sup\u003e ions were sparsely distributed, creating vacant pathways conducive to ion transport (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed).\u003c/p\u003e\u003cp\u003eThe intercalation of Cs⁺ ions modulated the membrane\u0026rsquo;s physical properties. They partially neutralized the negative surface charges, evidenced by a less negative zeta potential (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee), and consequently rendering the membranes less hydrophilic (Fig. S9). This would impart the membrane with higher anti-swelling capability, as will be demonstrated later.\u003c/p\u003e\u003cp\u003eFurthermore, deep dehydration promoted direct and tight binding between Cs⁺ ions and the vermiculite layers, potentially strengthening interlayer cohesion\u0026mdash;another favorable trait for swelling suppression. Indeed, X-ray photoelectron spectra (XPS) of dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM shows an increased binding energy of Cs4d and a new split peak of O1s, indicating stronger interaction between Cs\u003csup\u003e+\u003c/sup\u003e and vermiculite oxygen atoms\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef, Fig. S10). Nanoindentation tests further revealed substantially higher Young\u0026rsquo;s modulus in dp-Cs⁺-VM compared to Cs⁺-VM (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eg, Fig. S11), also suggesting the higher interlayer cohesion. This should also substantially enhance the anti-swelling capability.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eEnhanced anti-swelling capability\u003c/h3\u003e\n\u003cp\u003eTo assess the anti-swelling capability, we completely immersed the membranes in pure water. The air-dried neat VM quickly disintegrated in water (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). Then, the ion intercalation promoted the anti-swelling capability of X-VM, as suggested by the stable membrane form in water (Fig. S12-S14). However, they could not sustain rigorous mechanical disturbance as demonstrated by the poor stability under ultrasonication (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, Fig. S15). In contrast, all dp-X-VMs maintained their membrane form either in still water or under ultrasonication.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo better demonstrate the importance of the deep dehydration, we characterized the interlayer \u003cem\u003ed\u003c/em\u003e-spacings with X-ray diffraction (XRD) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec-d, Fig. S16). Without deep dehydration, the \u003cem\u003ed\u003c/em\u003e- spacing of the K\u003csup\u003e+\u003c/sup\u003e-VM swelled from 11.27 \u0026Aring; to 12.70 \u0026Aring;, and that of the Cs\u003csup\u003e+\u003c/sup\u003e-VM swelled from 11.30 \u0026Aring; to 12.45 \u0026Aring;, representing a swell ratio of 12.7% and 10.2%, respectively. For Mg\u003csup\u003e2+\u003c/sup\u003e-VM and Al\u003csup\u003e3+\u003c/sup\u003e-VM, the swelling ratio was 23.8% and 7.4%, respectively. After deep dehydration, the swelling was suppressed for all these membranes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). The dp-K\u003csup\u003e+\u003c/sup\u003e-VM swelled by 5.2% in water, while the dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM swelled even less, by merely 3.9% (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). On the other hand, the swelling ratio of the dp-Mg\u003csup\u003e2+\u003c/sup\u003e-VM was 12.6% and that of dp-Al\u003csup\u003e3+\u003c/sup\u003e-VM was 15.7%.\u003c/p\u003e\u003cp\u003eThese results also demonstrate that deep dehydrated membranes intercalated with monovalent ions have higher anti-swelling capability than those with multi-valent ions (Fig. S17). This contrasts with the trend that without deep dehydration, membranes intercalated by multivalent ions were much more stable in water. We note that the sequence of the swelling ratio (Al\u003csup\u003e3+\u003c/sup\u003e \u0026gt;Mg\u003csup\u003e2+\u003c/sup\u003e \u0026gt;K\u003csup\u003e+\u003c/sup\u003e \u0026gt;Cs\u003csup\u003e+\u003c/sup\u003e) coincides with that of the hydration enthalpy of these ions, suggesting that the hydration governs the swelling process. This aligns with prior studies indicating that clay swelling is initiated by the hydration of interlayer cations\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eTo better elucidate the different role of monovalent and multi-valent ions, we conducted DFT calculations. Models were constructed according to the measured \u003cem\u003ed\u003c/em\u003e-spacings after deep dehydration. For monovalent ions, the binding energy between the intercalated ion and the VM channel exceeded the ionic hydration energy in bulk water\u003csup\u003e\u003cspan additionalcitationids=\"CR32\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e(Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ef, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), suggesting that rehydration in dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM and dp-K\u003csup\u003e+\u003c/sup\u003e-VM is energetically unfavorable, leading to their high anti-swelling performance. For multivalent ions, however, the opposite trend was found. These ions have substantially higher hydration energy, exceeding the binding energy, consequently resulting in significant swelling in water.\u003c/p\u003e\u003cp\u003eAnother key question is why deep dehydration is essential for the anti-swelling capability. As suggested by Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, the deep dehydration process eliminates the interlayer water molecules, thereby removing hydration forces that hinders the collapse of the interlayer channels. The reduction in interlayer free spacing should subsequently enhances interlayer cohesion, since both the attractive van der Walls interaction and the attractive cation-vermiculite interaction increases with the decrease of the interaction distance, rendering the membrane less vulnerable to swelling. Indeed, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg, the calculated binding energy between Cs\u003csup\u003e+\u003c/sup\u003e ion and the vermiculite channel increases steadily as interlayer free spacing decreases (Table S2).\u003c/p\u003e\u003cp\u003eOverall, these results suggest that the interplay between the interlayer free spacing and the hydration energy of cations crucially determines the anti-swelling capability of the membrane. The deep dehydration process reduces the interlayer free spacing and thus enhances the binding energy, to an extent that surpasses the ionic hydration energy for membranes intercalated by monovalent cations. This prevents the rehydration of the interlayer cations and renders the membrane anti-swelling.\u003c/p\u003e\n\u003ch3\u003eEnhanced selectivity\u003c/h3\u003e\n\u003cp\u003eThe above results also suggest that dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM has the lowest swelling ratio and a narrow interlayer free spacing, making it promising for ion sieving applications. We then explored the Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e separation performance by experiments, DFT calculations, and \u003cem\u003eab initio\u003c/em\u003e molecular dynamics (AIMD) simulations. We chose Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e because their fast and high-selectivity separation is crucial but remains highly challenging for lithium extraction, especially from salt lake.\u003c/p\u003e\u003cp\u003eThe deep dehydration results in extremely narrow interlayer channels in water. This would significantly promote the steric hinderance against the transport of Mg\u003csup\u003e2+\u003c/sup\u003e ions and thus the Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity. In our AIMD simulations, we found that Mg\u003csup\u003e2+\u003c/sup\u003e could not enter the channels of the dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM in water, while Li\u003csup\u003e+\u003c/sup\u003e spontaneously entered (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea-b, Fig. S18). Indeed, for Mg\u003csup\u003e2+\u003c/sup\u003e, the first hydration shell must shed away two water molecules to enter the VM channel, whereas the hydration shell of Li\u003csup\u003e+\u003c/sup\u003e only undergoes deformation with virtually no change in the coordination number (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec-d).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn reality, some Mg\u003csup\u003e2+\u003c/sup\u003e ions would nevertheless permeate into the membrane through structural defects. Inside the VM channel, the Mg\u003csup\u003e2+\u003c/sup\u003e still experiences much larger energy barrier than Li\u003csup\u003e+\u003c/sup\u003e does due to the stronger binding between Mg\u003csup\u003e2+\u003c/sup\u003e and the VM, as suggested by our DFT calculations (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, Fig. S19, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). This guarantees that any Mg\u003csup\u003e2+\u003c/sup\u003e ion inside the membrane can still be hindered.\u003c/p\u003e\u003cp\u003eThe deep dehydration membranes were intercalated with many cations. This raises two questions\u0026mdash; Will the intercalated cations leak out? How do they affect the transport of Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e? First, as demonstrated by the AIMD simulations, the Cs\u003csup\u003e+\u003c/sup\u003e ion was trapped and could not leave the minimum-energy adsorption site as shown in the Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea. Second, with Cs\u003csup\u003e+\u003c/sup\u003e intercalation, we found that the mean square displacement (MSD) of the Li\u003csup\u003e+\u003c/sup\u003e increased rapidly with time while that of Mg\u003csup\u003e2+\u003c/sup\u003e increased slowly when both ions were allowed to freely migrate in the VM channel (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). In fact, when the Cs\u003csup\u003e+\u003c/sup\u003e fills the minimum-energy trap, Li\u003csup\u003e+\u003c/sup\u003e and Mg\u003csup\u003e2+\u003c/sup\u003e can only adsorb on higher-energy sites, leading to weakened ion-vermiculite binding and thus lower transport energy barrier (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eg, Fig. S20-S21). Accordingly, we found that Li\u003csup\u003e+\u003c/sup\u003e ions experienced much lower energy barrier (denoted by O-Al-O/Cs\u003csup\u003e+\u003c/sup\u003e) compared to that before Cs\u003csup\u003e+\u003c/sup\u003e intercalation (denoted by O-Al-O) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eh, Table S3). This could further enhance the Li\u003csup\u003e+\u003c/sup\u003e selectivity of dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM.\u003c/p\u003e\u003cp\u003eAltogether, our calculations and simulations suggest that the deep dehydrated membranes should exhibit high Li\u003csup\u003e+\u003c/sup\u003e permeability due to the reduced energy barrier after Cs\u003csup\u003e+\u003c/sup\u003e intercalation, and high Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity due to the narrow interlayer free spacing and the higher Mg\u003csup\u003e2+\u003c/sup\u003e-VM binding energy.\u003c/p\u003e\n\u003ch3\u003eLi/Mg separation for lithium extraction from salt lake\u003c/h3\u003e\n\u003cp\u003eEncouraged by the above results, we then evaluated the Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e separation performance of the membranes under concentration and electrical gradients, using a set of homemade dialysis or electrodialysis devices\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, Fig. S22). We first used a binary solution containing 0.1M LiCl and 0.1 M MgCl\u003csub\u003e2\u003c/sub\u003e for concentration-driven permeation test. The receiving solution was ultrapure water. In this experiment, the neat VM could not withstand prolonged tests due to its low stability in water. Following Cs\u003csup\u003e+\u003c/sup\u003e ion intercalation, the Cs\u003csup\u003e+\u003c/sup\u003e-VM exhibited moderate Li\u003csup\u003e+\u003c/sup\u003e selectivity. After deep dehydration, the Li\u003csup\u003e+\u003c/sup\u003e selectivity was substantially enhanced (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb, Fig. S23). For dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM, we confirmed that the optimal membrane thickness was ~\u0026thinsp;4 \u0026micro;m (Fig. S24). With such thickness, the Li\u003csup\u003e+\u003c/sup\u003e permeability was 0.49 mol m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and the Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity reached 92.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe are aware that in practical applications, external driving force, e.g., electric voltage or pressure, is often applied to increase the permeability. Our membrane does not allow the fast transport of water under pressure (Fig. S25, Table S4), and is therefore not suitable for pressure-driven nanofiltration. However, this is just highly demanded for electrodialysis since it prevents the backflow of water and ions carried by the water flow from low concentration side to the high concentration side (Fig. S26). By tuning the externally applied voltage from 0 to 1V, we found that a moderate voltage (0.1 V) could enhance the Li\u003csup\u003e+\u003c/sup\u003e permeability and thus the Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec, Fig. S27). Further increasing the voltage led to the decrease of Li\u003csup\u003e+\u003c/sup\u003e selectivity, presumably because such voltage is sufficient to force the divalent Mg\u003csup\u003e2+\u003c/sup\u003e entering the narrow channels and desorb from the vermiculite surfaces. With the optimal thickness and voltage, our membrane exhibits both high Li\u003csup\u003e+\u003c/sup\u003e permeability (0.65 mol m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and Mg\u003csup\u003e2+\u003c/sup\u003e selectivity (148), which clearly stand out among reported membranes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed, Table S5).\u003c/p\u003e\u003cp\u003eIn practice, the Mg\u0026sup2;⁺/Li⁺ molar ratio of the salt lake water vary significantly. Our membrane maintained relatively high selectivity when the Mg\u0026sup2;⁺/Li⁺ ratio increased from 2:1 to 100:1 (Fig. S28). We also varied the pH value of the feed solutions from 4 to 10 to account for the fact that most salt-lake water are weakly acidic to weakly alkaline\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. We found that the selectivity still kept above 80 (Fig. S29). Perhaps more interestingly, our membrane demonstrated robust acid and alkali resistance under these conditions (Fig. S30-S31), contrary to neat VM which is known to be highly vulnerable to acid and alkali. This should also be attributed to the fact that the Cs\u003csup\u003e+\u003c/sup\u003e ions are trapped on the VM surfaces, shielding them from being attacked.\u003c/p\u003e\u003cp\u003eRemarkably, thanks to the strong anti-swelling capability, our membrane exhibited robust Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e separation performance during long time (10 days) operation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee), which is crucial for industrial applications but remains rarely reported for previous high-selectivity Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e separation membranes.\u003c/p\u003e\u003cp\u003eThe high performance and high stability of the membrane allows potential practical lithium extraction applications. We used bittern from Qinghai salt lake (after solar evaporation, compositions shown in Table S6) and processed it with our membrane (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef, Fig. S32). By employing a one-step electrodialysis, the Mg\u003csup\u003e2+\u003c/sup\u003e/Li\u003csup\u003e+\u003c/sup\u003e ratio reduced from 15.1 to 0.5, and the content of the Li\u003csup\u003e+\u003c/sup\u003e increased from 6.1 wt% to 45.8 wt% (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg, Table S6). Then we concentrated the brines by 7.5 times, followed by precipitating the lithium with saturated Na\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e37,38\u003c/sup\u003e. The final product (white powders) has a Li\u003csup\u003e+\u003c/sup\u003e purity of 99.2% (Fig. S33). XRD patterns also confirmed the obtained high purity Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eh). The results demonstrate the great potential of dp-Cs\u003csup\u003e+\u003c/sup\u003e-VM membrane for highly efficient extraction of Li\u003csup\u003e+\u003c/sup\u003e from natural salt lake water.\u003c/p\u003e\u003cp\u003eBeyond its separation performance, the membrane\u0026rsquo;s minimal water crossover under high osmotic gradients reduces operational freshwater consumption. In addition, the cost the membrane is also low thanks to the cheap raw material vermiculite and the minimal usage of CsCl (~\u0026thinsp;1.28 g/m\u0026sup2;; Fig. S34, Table S7). This is crucial for reducing the cost of extracting lithium resources from salt lakes, which are typically located in arid regions.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe swelling of 2D layered membranes in aqueous environments has long hindered their practical application in separation processes. By exploiting the fact that the attractive van der Waals and electrostatic interaction increases with the decrease of interaction distance, we demonstrated a deep dehydration strategy to tackle this challenge, which forces the interlayer distance crossing a critical threshold that energetically disfavors rehydration and swelling. The deep dehydrated 2D layered vermiculite membranes exhibited high anti-swelling capability and narrow interlayer spacing. These enabled high Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity in long-time tests, allowing us to produce industrial-grade Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e from salt lake bittern through the electrodialysis\u0026ndash;precipitation process. Our current research focuses on the effect of deep dehydration process on membrane structure and properties, and the enhanced ion selective transport mechanism, which should shed light on design of other advanced ion sieving materials.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\n \u003ch2\u003eFabrication of the vermiculite nanosheets\u003c/h2\u003e\n \u003cp\u003eVermiculite nanosheets were prepared by delamination method we reported before\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. Typically, 10 g of thermally expanded vermiculite particles (2\u0026ndash;3 mm, purchased from Sigma-Aldrich) were dispersed in 150 mL of saturated NaCl solution and magnetically stirred for 24 h at 110\u0026deg;C for sufficient cation exchange. After repeated centrifugation washing with ethanol and water, the sediment was dispersed in 150 mL of 2 M LiCl solution and magnetically stirred for 24 h at 110\u0026deg;C. The Li\u003csup\u003e+\u003c/sup\u003e-exchanged vermiculite was then obtained after the same washing procedure as mentioned above. Then, the vermiculite slurry was mixed with 150 mL of 30% H\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e2\u003c/sub\u003e solution and magnetically stirred at 110\u0026deg;C for 24 h. The mixture was then ultrasonically treated 1 h for further exfoliation, followed by centrifugation at 6000 rpm for 10 min, to obtain the colloidal dispersion of monolayered nanosheets. The total mass of the resulting nanosheets is about 700 to 1000 mg.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eFabrication of the X-VM and dp-X-VM membranes\u003c/h3\u003e\n\u003cp\u003eThe lamellar VM membranes were assembled by the nanosheets via vacuum-filtrating. Typically, a certain amount of VM nanosheets were dispersed in at least 100 mL ultrapure water, then filtrated onto a PP substrate until the water was drained out. Afterwards, the VM membrane with PP substrate was heated at 60\u0026deg;C for 8 h, then free-standing VM membrane could be peeled from the substrate. For ion intercalation, 50 mL of 10\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e M CsCl (or KCl, MgCl\u003csub\u003e2\u003c/sub\u003e, and AlCl\u003csub\u003e3\u003c/sub\u003e) solution was added to the surface of a freshly assembled VM membrane. The filtration process was then continued until there is no visible water staining on the membrane surface. Given the low consumption of intercalation cations (Fig. S34, Table S7), the chloride salt solution could be collected to reuse.\u003c/p\u003e\n\u003cp\u003eThe deep dehydration process of VM membranes was enabled by programmed calcination in tube furnace (TL-1200, Nanjing Boyuntong Instruments, China) under inert atmosphere. The heat up and cool down rates are set to 10\u0026deg;C/min. Holding for 1 h after heating up to the specified temperature to ensure uniform dehydration at that temperature. The argon gas flow rate through the furnace was \u003cem\u003eca.\u003c/em\u003e 300 cm\u003csup\u003e3\u003c/sup\u003e min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e.\u003c/p\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eCharacterization\u003c/h2\u003e\n \u003cp\u003eSurface and cross-sectional morphology characterizations were carried out with a field-emission scanning electron microscopy (Hitachi SU-8010, Japan) integrated with energy-dispersive X-ray spectroscopy (EDS) and atomic force microscopy (AFM, Agilent 5400, USA). The lattice structure and atomic arrangement in interlayer spacings were obtained by scanning transmission election microscope (STEM, Thermo Scientific Themis Z, USA) with high-angle annular dark-field (HAADF). The surface charge property of the nanosheets and membrane samples was analyzed by zeta potential analyzer (Malvern Zetasizer Nano ZS90, UK) and electrokinetic analyzer (Anton Paar SurPASS3, Austria) respectively. The surface chemical properties were characterized by Fourier transform infrared spectrometry (FT-IR, Nicolet 6700, USA) and X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha, USA). The dehydration process of the materials was characterized by thermogravimetric analysis and Fourier-transform infrared spectroscopy coupled technology (TGA-FTIR, Netzsch STA-2500, Germany and Thermo Fisher IS-50, USA). XRD patterns were measured using X-ray diffractometer (Rigaku SmartLab 9KW, Japan). The water contact angle was measured using a contact angle goniometer (LAUDA Scientific LSA 100, Germany). The mechanical properties are tested with nano indenter (Bruker Hysitron TI 980, Germany) for three different area of one sample.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eIon selectivity and lithium extraction\u003c/h2\u003e\n \u003cp\u003eThe diffusion dialysis experiments were realized by the self-made permeation apparatus with two polytetrafluoroethylene (PTFE) reservoirs. The tested membrane (effective area 7.07 mm\u003csup\u003e2\u003c/sup\u003e) was mounted in the center to separate two sides. During the experiments, one reservoir was filled with 10 ml feed solution containing LiCl and MgCl\u003csub\u003e2\u003c/sub\u003e (0.1 M unless otherwise specified) while 10 ml ultrapure water was added in the permeation reservoir. For pH-stability permeation tests, trace amounts of NaOH or HCl solution were used to adjust the pH of the feed solution. For the external electric field driven permeation process, inert graphite plate electrodes (30\u0026times;10\u0026times;1 mm) were introduced in the reservoir, and a DC power supply was used to control the bias voltage between the two sides. The electrode in the high concentration side was connected to the positive pole of the power supply.\u003c/p\u003e\n \u003cp\u003eThe salt lake bittern was obtained from Qinghai Lake, China (Table S6). The continuous electrodialysis was performed using homemade setups (Fig. S32). The lithium-rich solution after electrodialysis was further treated to obtain Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e product\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. In general, a saturated Na\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e solution was added to the collected lithium-rich solution at 80\u0026deg;C\u003csup\u003e38\u003c/sup\u003e. The obtained white sediments can be further refined by dissolving in ultrapure water under 20\u0026deg;C and then heated to 90\u0026deg;C to filter the precipitate while hot. The generated Li\u003csub\u003e2\u003c/sub\u003eCO\u003csub\u003e3\u003c/sub\u003e were collected and washed by centrifuge and then dried in an air-forced oven at 70\u0026deg;C for at least 12 h. The sediments were dissolved in dilute hydrochloric acid solutions to determine the lithium purity.\u003c/p\u003e\n \u003cp\u003eThe concentration of cations in the permeated solution and the salt lake water were measured by inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7850, USA) for ppb level of concentration and inductively coupled plasma atomic emission spectroscopy (ICP-OES, Agilent 5800, USA) for ppm level of concentration. The anion permeation concentration was measured by ion chromatograph (Metrohm 863 Basic IC plus, Switzerland). At least three parallel samples were tested to evaluate systematic errors. The Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity in the solution was calculated with molar concentrations by the following Eq. (\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e),\u003c/p\u003e\n \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e$$\\:{S}_{{\\text{L}\\text{i}}^{+}/{\\text{M}\\text{g}}^{2+}}\\:=\\:\\frac{{\\left({C}_{{\\text{L}\\text{i}}^{+}}/{C}_{{\\text{M}\\text{g}}^{2+}}\\right)}_{dialysate}}{{\\left({{C}_{{\\text{L}\\text{i}}^{+}}/C}_{{\\text{M}\\text{g}}^{2+}}\\right)}_{feed\\:solution}}$$\u003c/div\u003e\n \u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eDensity functional theory (DFT) calculation\u003c/h2\u003e\n \u003cp\u003eThe DFT calculations of non-periodic cluster models were performed on ORCA 6.0.0 program packages\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. The optimized structures were conducted at B3LYP level of theory with def2-TZVP basis set\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. All calculations have contained the Grimme\u0026apos;s D3 dispersion correction (BJ-damping) and taken the solvent effect into account\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Periodic DFT calculations were conducted on CP2K 2024.1 package\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. The VM single unit cell (lattice parameters a\u0026thinsp;=\u0026thinsp;5.349 \u0026Aring;, b\u0026thinsp;=\u0026thinsp;9.255 \u0026Aring;, c\u0026thinsp;=\u0026thinsp;28.89 \u0026Aring; and \u0026alpha;\u0026thinsp;=\u0026thinsp;\u0026gamma;\u0026thinsp;=\u0026thinsp;90.00\u0026deg;, \u0026beta;\u0026thinsp;=\u0026thinsp;97.12\u0026deg;) was expanded by replication into a 2\u0026times;2\u0026times;1 supercell with interlayer-filled magnesium ions removed. The number of water molecules filled at different interlayer \u003cem\u003ed\u003c/em\u003e-spacings was referenced from literature reports\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. The vacuum layer thickness is large enough to avoid any possible interaction between the repeated slabs. Perdew-Burke-Ernzerhof (PBE) with Grimme\u0026rsquo;s D3 dispersion corrections were employed for the exchange and correlation functional\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. The core electrons were treated using the Goedecker-Tetter-Hutter (GTH) pseudo-potentials, and the valence electrons were expanded using the Gaussian and Plane-Wave (GPW) combined basis sets (DZVP-MOLOPT for structural optimization and vibrational analysis, TZVP for energy calculation with 400 Ry cut-off energy) \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. The Monkhorst-Pack method was used to generate the Brillouin zone integration from a Gamma centered 4\u0026times;2\u0026times;1 k-point mesh. All wave function analyses (including non-covalent interaction using independent gradient model based on Hirshfeld partition\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, Hirshfeld-I atomic charges\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e) and visualizations were done with the help of Multiwfn (3.8dev)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e and VMD (1.9.3) software\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003eThe energy barrier for a single ion transporting through the different sites of the channel using the climbing image nudged elastic band (CI-NEB) method\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e. All binding energy calculations take the basis set superposition error (BSSE) into consideration\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e, thus the interaction of a given ion (M\u003csup\u003en+\u003c/sup\u003e) with VM nanosheet is calculated as Eq.\u0026nbsp;(\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e),\u003c/p\u003e\n \u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e$$\\:{E}_{Binding}={E}_{{M}^{n+}@VM}-{E}_{{M}^{n+}}-{E}_{VM}+{E}_{BSSE}$$\u003c/div\u003e\n \u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eThe Shermo program was used to calculate thermodynamic data, the system temperature and pressure were set to 298.15 K and 1 atm respectively\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eAb-initio\u003c/strong\u003e \u003cstrong\u003emolecular dynamics (AIMD) simulation\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003eAIMD calculations were performed to study the structure and mobility of hydrated ions in VM nanochannel using the same theoretical level of DFT calculations with CP2K. The temperature was fixed at 298.15 K using the Nos\u0026eacute;-Hoover thermostat with a damping constant of 100 fs. Periodic boundary conditions were applied in all directions.\u003c/p\u003e\n \u003cp\u003eThe mean square displacement (MSD) was calculated by Eq.\u0026nbsp;(\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e),\u003c/p\u003e\n \u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e$$\\:MSD=\\frac{1}{N}\\sum\\:_{i=1}^{N}{\\left[{r}_{i}\\left(t\\right)-{r}_{i}\\left(0\\right)\\right]}^{2}$$\u003c/div\u003e\n \u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:N\\)\u003c/span\u003e\u003c/span\u003e is the total number of observed species in the unit cell, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{r}_{i}\\left(t\\right)\\)\u003c/span\u003e\u003c/span\u003e is the position of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:i\\)\u003c/span\u003e\u003c/span\u003e-th species at time \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003eThen the self-diffusion coefficient (\u003cem\u003eD\u003c/em\u003e) of the cation was fitted from the slope of MSD curves as Eq.\u0026nbsp;(\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e),\u003c/p\u003e\n \u003cdiv id=\"Equ4\" class=\"Equation\"\u003e\n \u003cdiv class=\"mathdisplay\" id=\"FileID_Equ4\" name=\"EquationSource\"\u003e$$\\:D=\\frac{1}{6}\\underset{t\\to\\:\\infty\\:}{\\text{lim}}\\frac{MSD\\left(t\\right)}{t}$$\u003c/div\u003e\n \u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eFor the non-equilibrium simulations of the entrance of ions into the nanochannels, a small velocity (0.002 \u0026Aring; fs\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) was used on cations to perform constrained AIMD, which could mimic the migration of ions under a concentration gradient. The nanosheet was fixed throughout the simulation in order to collect the trajectories of specific ions.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data supporting the findings of this study are available within the paper and its Supplementary Information. Source data are provided with this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the\u0026nbsp;National Key R\u0026amp;D Program of China (NO. 2022YFB3805900), the National Natural Science Foundation of China (22272194), Shandong Provincial Natural Science Foundation (ZR2021YQ12, ZR2024QB294), QIBEBT/SEI/QNESL (S202303), Key R\u0026amp;D Program of Shandong Province (2022CXGC010302) and the China Postdoctoral Science Foundation (No. 2024M753353). The authors acknowledge the use of DeepSeek v3.1 to correct language errors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eZ.M. conducted the experiments and simulations with help from J.F., C.Z. and H.G., Z.M. wrote and revised the draft. J.G. conceptualized and supervised the project and revised the draft.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interest\u003c/strong\u003e\u003cstrong\u003es\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary information\u0026nbsp;\u003c/strong\u003eis available online.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrespondence and requests for materials\u003c/strong\u003e should be addressed to J. G.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eColeman, J. N. \u003cem\u003eet al.\u003c/em\u003e Two-Dimensional Nanosheets Produced by Liquid Exfoliation of Layered Materials. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e331\u003c/strong\u003e, 568\u0026ndash;571 (2011).\u003c/li\u003e\n\u003cli\u003eGao, J., Feng, Y., Guo, W. \u0026amp; Jiang, L. Nanofluidics in two-dimensional layered materials: inspirations from nature. \u003cem\u003eChem. Soc. Rev.\u003c/em\u003e \u003cstrong\u003e46\u003c/strong\u003e, 5400\u0026ndash;5424 (2017).\u003c/li\u003e\n\u003cli\u003eCai, X., Luo, Y., Liu, B. \u0026amp; Cheng, H.-M. Preparation of 2D material dispersions and their applications. \u003cem\u003eChem. Soc. Rev.\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 6224\u0026ndash;6266 (2018).\u003c/li\u003e\n\u003cli\u003eKang, Y., Xia, Y., Wang, H. \u0026amp; Zhang, X. 2D Laminar Membranes for Selective Water and Ion Transport. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1902014 (2019).\u003c/li\u003e\n\u003cli\u003eHuang, T. \u003cem\u003eet al.\u003c/em\u003e Two-Dimensional Sodium Channels with High Selectivity and Conductivity for Osmotic Power Generation from Wastewater. \u003cem\u003eACS Nano\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 17245\u0026ndash;17253 (2023).\u003c/li\u003e\n\u003cli\u003eYang, R. \u003cem\u003eet al.\u003c/em\u003e Synthesis of atomically thin sheets by the intercalation-based exfoliation of layered materials. \u003cem\u003eNat. Synth.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 101\u0026ndash;118 (2023).\u003c/li\u003e\n\u003cli\u003eYeh, C.-N., Raidongia, K., Shao, J., Yang, Q.-H. \u0026amp; Huang, J. On the origin of the stability of graphene oxide membranes in water. \u003cem\u003eNat. Chem.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 166\u0026ndash;170 (2015).\u003c/li\u003e\n\u003cli\u003eZheng, S., Tu, Q., Urban, J. J., Li, S. \u0026amp; Mi, B. Swelling of Graphene Oxide Membranes in Aqueous Solution: Characterization of Interlayer Spacing and Insight into Water Transport Mechanisms. \u003cem\u003eACS Nano\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 6440\u0026ndash;6450 (2017).\u003c/li\u003e\n\u003cli\u003eLiu, Y. \u003cem\u003eet al.\u003c/em\u003e Pillared Laminar Vermiculite Membranes with Tunable Monovalent and Multivalent Ion Selectivity. \u003cem\u003eAdv. Mater.\u003c/em\u003e 2417994 (2025) doi:10.1002/adma.202417994.\u003c/li\u003e\n\u003cli\u003eLu, Z. \u003cem\u003eet al.\u003c/em\u003e Self-Crosslinked MXene (Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003cem\u003e\u003csub\u003ex\u003c/sub\u003e\u003c/em\u003e) Membranes with Good Antiswelling Property for Monovalent Metal Ion Exclusion. \u003cem\u003eACS Nano\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 10535\u0026ndash;10544 (2019).\u003c/li\u003e\n\u003cli\u003eDing, L. \u003cem\u003eet al.\u003c/em\u003e Effective ion sieving with Ti3C2Tx MXene membranes for production of drinking water from seawater. \u003cem\u003eNat. Sustain.\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 296\u0026ndash;302 (2020).\u003c/li\u003e\n\u003cli\u003eHelal, M. I. \u003cem\u003eet al.\u003c/em\u003e Understanding the swelling behavior of Ti\u003csub\u003e3\u003c/sub\u003eC\u003csub\u003e2\u003c/sub\u003eT\u003cem\u003e\u003csub\u003ex\u003c/sub\u003e\u003c/em\u003e MXene membranes in aqueous media. \u003cem\u003eJ. Mater. Chem. A\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 30729\u0026ndash;30742 (2024).\u003c/li\u003e\n\u003cli\u003eChen, L. \u003cem\u003eet al.\u003c/em\u003e Ion sieving in graphene oxide membranes via cationic control of interlayer spacing. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e550\u003c/strong\u003e, 380\u0026ndash;383 (2017).\u003c/li\u003e\n\u003cli\u003eWang, W. \u003cem\u003eet al.\u003c/em\u003e 2D Vermiculite Nanolaminated Membranes for Efficient Organic Solvent Nanofiltration. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e 2410635 (2024) doi:10.1002/adfm.202410635.\u003c/li\u003e\n\u003cli\u003eHribar, B., Southall, N. T., Vlachy, V. \u0026amp; Dill, K. A. How ions affect the structure of water. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e124\u003c/strong\u003e, 12302\u0026ndash;12311 (2002).\u003c/li\u003e\n\u003cli\u003eLeite, F. L., Bueno, C. C., Da R\u0026oacute;z, A. L., Ziemath, E. C. \u0026amp; Oliveira, O. N. Theoretical models for surface forces and adhesion and their measurement using atomic force microscopy. \u003cem\u003eInt. J. Mol. Sci.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 12773\u0026ndash;12856 (2012).\u003c/li\u003e\n\u003cli\u003eDuong, D. L., Yun, S. J. \u0026amp; Lee, Y. H. van der waals layered materials: Opportunities and challenges. \u003cem\u003eACS Nano\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 11803\u0026ndash;11830 (2017).\u003c/li\u003e\n\u003cli\u003eVasu, K. S. \u003cem\u003eet al.\u003c/em\u003e Van der waals pressure and its effect on trapped interlayer molecules. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, (2016).\u003c/li\u003e\n\u003cli\u003eMotokawa, R. \u003cem\u003eet al.\u003c/em\u003e Collective Structural Changes in Vermiculite Clay Suspensions Induced by Cesium Ions. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 6585 (2014).\u003c/li\u003e\n\u003cli\u003eYin, X., Takahashi, H., Inaba, Y. \u0026amp; Takeshita, K. Desorption of Cesium Ions from Vermiculite with Sea Water by Hydrothermal Process. \u003cem\u003eChem. Lett.\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 256\u0026ndash;258 (2016).\u003c/li\u003e\n\u003cli\u003eRazmjou, A., Asadnia, M., Hosseini, E., Habibnejad Korayem, A. \u0026amp; Chen, V. Design principles of ion selective nanostructured membranes for the extraction of lithium ions. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 5793 (2019).\u003c/li\u003e\n\u003cli\u003eHou, L. \u003cem\u003eet al.\u003c/em\u003e Understanding the Ion Transport Behavior across Nanofluidic Membranes in Response to the Charge Variations. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 2009970 (2021).\u003c/li\u003e\n\u003cli\u003eZhang, Y. \u003cem\u003eet al.\u003c/em\u003e Congener-welded crystalline carbon nitride membrane for robust and highly selective Li/Mg separation. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, eadm9620 (2024).\u003c/li\u003e\n\u003cli\u003eLi, Z. \u003cem\u003eet al.\u003c/em\u003e Lithium extraction from brine through a decoupled and membrane-free electrochemical cell design. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e385\u003c/strong\u003e, 1438\u0026ndash;1444 (2024).\u003c/li\u003e\n\u003cli\u003eYang, S., Wang, Y., Pan, H., He, P. \u0026amp; Zhou, H. Lithium extraction from low-quality brines. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e636\u003c/strong\u003e, 309\u0026ndash;321 (2024).\u003c/li\u003e\n\u003cli\u003eFernandez, R., Martirena, F. \u0026amp; Scrivener, K. L. The origin of the pozzolanic activity of calcined clay minerals: A comparison between kaolinite, illite and montmorillonite. \u003cem\u003eCem. Concr. Res.\u003c/em\u003e \u003cstrong\u003e41\u003c/strong\u003e, 113\u0026ndash;122 (2011).\u003c/li\u003e\n\u003cli\u003eXie, W. \u003cem\u003eet al.\u003c/em\u003e Evolution of the crystallographic structure and physicochemical aspects of rectorite upon calcination. \u003cem\u003eAppl. Clay Sci.\u003c/em\u003e \u003cstrong\u003e185\u003c/strong\u003e, 105374 (2020).\u003c/li\u003e\n\u003cli\u003eDupin, J.-C., Gonbeau, D., Vinatier, P. \u0026amp; Levasseur, A. Systematic XPS studies of metal oxides, hydroxides and peroxides. \u003cem\u003ePhys. Chem. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1319\u0026ndash;1324 (2000).\u003c/li\u003e\n\u003cli\u003eHensen, E. J. M. \u0026amp; Smit, B. Why clays swell. \u003cem\u003eJ. Phys. Chem. B\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 12664\u0026ndash;12667 (2002).\u003c/li\u003e\n\u003cli\u003eTeich-McGoldrick, S. L., Greathouse, J. A., Jov\u0026eacute;-Col\u0026oacute;n, C. F. \u0026amp; Cygan, R. T. Swelling properties of montmorillonite and beidellite clay minerals from molecular simulation: Comparison of temperature, interlayer cation, and charge location effects. \u003cem\u003eJ. Phys. Chem. C\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, 20880\u0026ndash;20891 (2015).\u003c/li\u003e\n\u003cli\u003eRandles, J. E. B. The real hydration energies of ions. \u003cem\u003eTrans. Faraday Soc.\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 1573\u0026ndash;1581 (1956).\u003c/li\u003e\n\u003cli\u003eKollman, P. A. \u0026amp; Kuntz, I. D. Studies of Cation hydration. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e94\u003c/strong\u003e, 9236\u0026ndash;9237 (1972).\u003c/li\u003e\n\u003cli\u003eKepp, K. P. Free energies of hydration for metal ions from heats of vaporization. \u003cem\u003eJ. Phys. Chem. A\u003c/em\u003e \u003cstrong\u003e123\u003c/strong\u003e, 6536\u0026ndash;6546 (2019).\u003c/li\u003e\n\u003cli\u003eBao, S. \u003cem\u003eet al.\u003c/em\u003e Randomly oriented covalent organic framework membrane for selective li+ sieving from other ions. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 3896 (2025).\u003c/li\u003e\n\u003cli\u003eZachara, J. M. \u003cem\u003eet al.\u003c/em\u003e Geo- and biogeochemical processes in a heliothermal hypersaline lake. \u003cem\u003eGeochim. Cosmochim. Acta\u003c/em\u003e \u003cstrong\u003e181\u003c/strong\u003e, 144\u0026ndash;163 (2016).\u003c/li\u003e\n\u003cli\u003eGhori, N.-U.-H., Wise, M. J. \u0026amp; Whiteley, A. S. Temporal Microbial Community Dynamics Within a Unique Acid Saline Lake. \u003cem\u003eFront. Microbiol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 649594 (2021).\u003c/li\u003e\n\u003cli\u003eLinneen, N., Bhave, R. \u0026amp; Woerner, D. Purification of industrial grade lithium chloride for the recovery of high purity battery grade lithium carbonate. \u003cem\u003eSep. Purif. Technol.\u003c/em\u003e \u003cstrong\u003e214\u003c/strong\u003e, 168\u0026ndash;173 (2019).\u003c/li\u003e\n\u003cli\u003eZhang, S. \u003cem\u003eet al.\u003c/em\u003e Solar-driven membrane separation for direct lithium extraction from artificial salt-lake brine. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 238 (2024).\u003c/li\u003e\n\u003cli\u003eShao, J.-J., Raidongia, K., Koltonow, A. R. \u0026amp; Huang, J. Self-assembled two-dimensional nanofluidic proton channels with high thermal stability. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 7602 (2015).\u003c/li\u003e\n\u003cli\u003eRettig, S. L., Jones, B. F. \u0026amp; Risacher, F. Geochemical evolution of brines in the Salar of Uyuni, Bolivia. \u003cem\u003eChem. Geol.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 57\u0026ndash;79 (1980).\u003c/li\u003e\n\u003cli\u003eAn, J. W. \u003cem\u003eet al.\u003c/em\u003e Recovery of lithium from Uyuni salar brine. \u003cem\u003eHydrometallurgy\u003c/em\u003e \u003cstrong\u003e117\u0026ndash;118\u003c/strong\u003e, 64\u0026ndash;70 (2012).\u003c/li\u003e\n\u003cli\u003eNeese, F. Software update: The ORCA program system\u0026mdash;Version 5.0. \u003cem\u003eWIREs Comput. Mol. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, e1606 (2022).\u003c/li\u003e\n\u003cli\u003eWeigend, F. \u0026amp; Ahlrichs, R. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. \u003cem\u003ePhys. Chem. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 3297 (2005).\u003c/li\u003e\n\u003cli\u003eGrimme, S., Antony, J., Ehrlich, S. \u0026amp; Krieg, H. A consistent and accurate \u003cem\u003eab initio\u003c/em\u003e parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e132\u003c/strong\u003e, 154104 (2010).\u003c/li\u003e\n\u003cli\u003eK\u0026uuml;hne, T. D. \u003cem\u003eet al.\u003c/em\u003e CP2K: An electronic structure and molecular dynamics software package - Quickstep: Efficient and accurate electronic structure calculations. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e152\u003c/strong\u003e, 194103 (2020).\u003c/li\u003e\n\u003cli\u003ePerdew, J. P., Burke, K. \u0026amp; Ernzerhof, M. Generalized Gradient Approximation Made Simple. \u003cem\u003ePhys. Rev. Lett.\u003c/em\u003e \u003cstrong\u003e77\u003c/strong\u003e, 3865\u0026ndash;3868 (1996).\u003c/li\u003e\n\u003cli\u003eGoedecker, S., Teter, M. \u0026amp; Hutter, J. Separable dual-space Gaussian pseudopotentials. \u003cem\u003ePhys. Rev. B\u003c/em\u003e \u003cstrong\u003e54\u003c/strong\u003e, 1703\u0026ndash;1710 (1996).\u003c/li\u003e\n\u003cli\u003eKrack, M. Pseudopotentials for H to Kr optimized for gradient-corrected exchange-correlation functionals. \u003cem\u003eTheor. Chem. Acc.\u003c/em\u003e \u003cstrong\u003e114\u003c/strong\u003e, 145\u0026ndash;152 (2005).\u003c/li\u003e\n\u003cli\u003eLu, T. \u0026amp; Chen, Q. Independent gradient model based on Hirshfeld partition: A new method for visual study of interactions in chemical systems. \u003cem\u003eJ. Comput. Chem.\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 539\u0026ndash;555 (2022).\u003c/li\u003e\n\u003cli\u003eBultinck, P., Van Alsenoy, C., Ayers, P. W. \u0026amp; Carb\u0026oacute;-Dorca, R. Critical analysis and extension of the Hirshfeld atoms in molecules. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 144111 (2007).\u003c/li\u003e\n\u003cli\u003eLu, T. \u0026amp; Chen, F. Multiwfn: A multifunctional wavefunction analyzer. \u003cem\u003eJ. Comput. Chem.\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, 580\u0026ndash;592 (2012).\u003c/li\u003e\n\u003cli\u003eHumphrey, W., Dalke, A. \u0026amp; Schulten, K. VMD: Visual molecular dynamics. \u003cem\u003eJ. Mol. Graph.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 33\u0026ndash;38 (1996).\u003c/li\u003e\n\u003cli\u003eHenkelman, G., Uberuaga, B. P. \u0026amp; J\u0026oacute;nsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. \u003cem\u003eJ. Chem. Phys.\u003c/em\u003e \u003cstrong\u003e113\u003c/strong\u003e, 9901\u0026ndash;9904 (2000).\u003c/li\u003e\n\u003cli\u003eKestner, N. R. \u0026amp; Combariza, J. E. Basis Set Superposition Errors: Theory and Practice. in \u003cem\u003eReviews in Computational Chemistry\u003c/em\u003e 99\u0026ndash;132 (1999). doi:10.1002/9780470125908.ch2.\u003c/li\u003e\n\u003cli\u003eLu, T. \u0026amp; Chen, Q. Shermo: A general code for calculating molecular thermochemistry properties. \u003cem\u003eComput. Theor. Chem.\u003c/em\u003e \u003cstrong\u003e1200\u003c/strong\u003e, 113249 (2021).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7726847/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7726847/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTwo-dimensional layered membranes show great promise for separation applications but long suffered from the swelling problem even when intercalated with binding agents such as cations. Swelling is initiated by hydration forces yet restrained by van der Waals attraction and interlayer-ion-bridged electrostatic interactions\u0026mdash;both of which increase sharply with the decrease of interaction distance. We therefore demonstrate a deep dehydration strategy to tackle the swelling problem. The deep dehydration at high temperature almost completely removes interlayer water molecules, eliminating hydration-induced repulsion and thus greatly compresses the interlayer spacings. This in turn substantially enhances interlayer attraction and makes rehydration energetically unfavorable. Our deep-dehydrated Cs\u003csup\u003e+\u003c/sup\u003e-intercalated vermiculite membrane maintains a narrow interlayer spacing of \u0026lt;\u0026thinsp;0.3 nm in water for over one month, with a swelling ratio of only 4%. These stable nanochannels achieve Li\u003csup\u003e+\u003c/sup\u003e/Mg\u003csup\u003e2+\u003c/sup\u003e selectivity of 148 and Li\u003csup\u003e+\u003c/sup\u003e permeability of 0.65 mol m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e h\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, ranking among the best reported performances, and enable production of industrial-grade lithium carbonate via electrodialysis\u0026ndash;precipitation. Our results highlight the critical influence of hydration/dehydration history on layered membrane swelling and performance.\u003c/p\u003e","manuscriptTitle":"Deep dehydrated layered membrane for selective and fast ion separation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-13 12:32:21","doi":"10.21203/rs.3.rs-7726847/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"67c81b43-8ad2-43f8-8b9e-ac00763750dc","owner":[],"postedDate":"October 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":56050493,"name":"Physical sciences/Materials science"},{"id":56050494,"name":"Physical sciences/Nanoscience and technology"}],"tags":[],"updatedAt":"2026-02-05T15:16:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-13 12:32:21","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7726847","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7726847","identity":"rs-7726847","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}