Intercalating cation specific self-repairing of vermiculite nanofluidic membrane

Raj Kumar Gogoi and Kalyan Raidongia *
Department of Chemistry, Indian Institute of Technology Guwahati, Guwahati, 781039, Assam, India. E-mail:

Received 27th February 2018 , Accepted 25th April 2018

First published on 26th April 2018

The balance between expanding and contracting forces of lamellar clay minerals was exploited for preparing self-repairable nanofluidic membranes. Application of a tiny drop of water (20 μL) not only healed physical damages of lamellar vermiculite membrane, such as punching a hole, scratching with a sharp object or breaking into multiple pieces, but also completely recovered its nanofluidic ionic conductivity. The extraordinary thermal stability of the clay layer was exploited to perform the healing experiment at elevated temperature, reducing the healing time to just 30 seconds at 140 °C. Experiments with different charge balancing cations and in situ monitoring of the healing process through X-ray diffraction studies suggest that cations with hydration energy higher than the energy of the clay-cation attraction, such as that of Li+, initiate water-assisted swelling of the layers, which leads to the re-assembling of the flakes in the damaged sites. The dependence of vermiculite-swelling on the nature of interlayer cations is also exploited for preparing highly responsive bilayer membranes by assembling clay flakes with different charge balancing cations. The water-assisted swelling of Li-vermiculite is also utilized to fuse them with lamellar membranes of other 2D materials such as graphene oxide and vanadium pentoxide. Even though vermiculite sheets are electrically insulators, air dried vermiculite membranes exhibit an interesting self-healable humidity dependent conductivity.


One of the most desired qualities in any technologically important functional material is long-time durability because the protection of delicate functional materials in a realistic environment is a very expensive and troublesome process. The long-time durability of materials arises either through their intrinsic robustness or through efficient damage controlling mechanisms. For example, many biological materials that do not possess intrinsic robustness exhibit excellent durability through effective damage controlling mechanisms, such as self-healing or assisted-healing of the damaged sites. Inspired by nature, several efforts have been devoted to imbue self-healing qualities into various man-made materials such as polymers,1–4 metals,5,6 ceramics and concrete.7–9 Several impetuses such as moisture,10–12 heat,8,10 light13–15 or pH16 have also been shown to induce healing of materials through reversible hydrogen bonding,1,2 π–π interactions,11 metal–ligand interaction,14,17 ionic interactions or by molecular rearrangement.10 Along with reinstating to original physical structures, many of these materials have also been shown to recover their functional properties such as stimuli responsiveness,11,16 self-cleaning,18 super-hydrophobicity,19–21 anti-corrosiveness22,23 adhesiveness24 and electrical conductivity.1,3,11 However, fabrication of materials with self-healable functionality is still a very challenging task as it requires precisely-controlled and minutely balanced chemical interactions. Long-time durability/self-healing is particularly challenging for materials that operate in conjunction with liquids such as nano/microfluidic devices, filtration membranes or proton conducting channels. Therefore, reports on self-healing nanofluidic devices or proton conducting membranes are very scarce in the literature. Herein, we utilized the minute balance between the oppositely acting forces of clay minerals, viz., contraction due to the attraction of clay surface towards charge balancing cation and expansion due to the hydration of interlayer cations to create a self-repairable nanofluidic system.

The emergence of size-dependent properties of fluids confined in the nanometer size-regime has drawn tremendous attention from scientific communities across disciplines. New properties and phenomena of liquids such as surface-charge-governed ionic conductivity,25,26 ionic current rectifications,27,28 and gated and selected ion/molecular transport29 have laid the platform for application of nanofluidic channels in diverse areas such as molecular sieving,30,31 energy harvesting through reverse electrodialysis or pressure retarded osmosis,32,33 manipulation of single molecules,34,35 and fuel cells and batteries.36 The primary requirement for all these applications is materials with a compact assembly of nanofluidic channels that possesses long-time durability in realistic environments. Studies on nanofluidic channels have also opened up a new arena for understanding the activity of biological nanochannels. A tremendous amount of research efforts have been devoted to imitating biological activities in artificial channels that mimic the properties of biological nanochannels. As most of the biological channels are self-healing in nature, a self-healable nanofluidic channel is essential to attain the representative conditions of biological systems. Herein, we demonstrate intercalating cation specific self-repairing of nanofluidic-channel-networks fabricated by restacking exfoliated layers of 2D materials. The percolated network of two-dimensional nanochannels prepared by self-assembling exfoliated layers of two-dimensional (2D) materials has already provided a new platform to study ionic/molecular transport under sub-nanometer confinement. By virtue of the ease of preparation and scalability along with the abundance of layered materials and reconstructed lamellar membranes, self-healing ability has promising practical applications in areas such as water desalination, molecular sieving, and energy storage and harvesting. Alongside these technologically important applications, induction of self-healing properties to the molecularly thin nanofluidic network of lamellar membranes could open-up new directions of nanofluidic study.

Results and discussions

The freestanding nanochannel-network prepared by self-assembling exfoliated layers of the Li-exchanged vermiculite clay was chosen herein as a representative system to explore the healing-capability of reconstructed lamellar membranes. Along with the excellent proton conductivity, vermiculite membranes exhibit very high thermal and chemical stability; moreover, they can be easily combined with exfoliated layers of other layered materials to prepare highly responsive smart materials.37,38 The molecularly thin exfoliated clay layers employed herein were obtained through two-step ion exchange (Na+ followed by Li+) reactions on thermally expanded vermiculite crystals. Representative microscopic images of the Li+ exchanged clay layers are shown in the ESI Fig. S1. Typically, lateral dimensions of the flakes vary in the range of 60–5 μm2 with an average height of ∼4 nm. A representative freestanding membrane prepared through vacuum-filtration-assisted self-assembly of the well-dispersed Li-exchanged vermiculite flakes is shown in Fig. 1a. The smooth outer-surface and lamellar cross-section of the membrane are clearly visible from the field emission scanning electron microscopy (FESEM) examination; representative images are shown in Fig. 1b and c, respectively. Fig. 1d compares XRD patterns of un-exfoliated vermiculite crystals with that of the Li-exchanged layers restacked into membrane form. Notably, the multiple XRD reflections (in the range of 2θ = 6.4 to 9 degree) of the initial vermiculite crystals, which are attributed to the presence of different interlayer cations and different degrees of hydration,39 converged into a single reflection in the re-stacked membrane, revealing successful replacement of the interlayer cations.
image file: c8ta01885e-f1.tif
Fig. 1 Characterization of Li-vermiculite membrane (a) digital photo of a freestanding membrane prepared by vacuum filtration of exfoliated layers of vermiculite clay charge balanced with Li+ ions. FESEM images of the (b) surface, and (c) cross-section of a Li-vermiculite membrane. (d) Comparison of XRD patterns of un-exfoliated vermiculite particles (black curve) with that of the reconstructed membrane (red curve).

In order to explore the healing ability, a Li-vermiculite membrane (thickness 40 μm) was punched with a sharp pin to create an artificial hole of dimensions ∼0.4 mm × 0.3 mm × 40 μm and exposed the same to a tiny droplet of water (20 μL) at room temperature. The hole on the surface of the membrane disappeared after 20 minutes. The event was monitored under an optical microscope, and the snapshots are shown in Fig. 2a. Similarly, the surface of a Li-vermiculite membrane was scratched with a sharp object to create an artificial crack of dimension ∼0.15 mm × 5 mm. A droplet of water (20 μL) was introduced to the damage-site at room temperature. As the water droplet started evaporating the crack site became blurry and disappeared after 20 minutes; snapshots are shown in ESI Fig. S2. This automatic disappearance of an artificial crack-site can be attributed to self-re-organization of the Li-exchanged flakes with the help of liquid water. Similarly, the FESEM images of a cut made at the edge of vermiculite strip and its subsequent repairing are shown in ESI Fig. S3. The healing ability was further confirmed by placing edges of two rectangular strips of the Li-exchanged membrane (dimension 35 mm × 6 mm × 0.035 mm) together, followed by addition of a water droplet (20 μL) to the junction. When the evaporation of water molecules under ambient conditions was completed, both of the strips were fused into a single piece. The healed strip was found to be as flexible as the pristine strip of similar dimensions. The healed strip was dragged again until its failure and remarkably, the strip broke in a new place preserving the freshly healed area. The snapshots of the healing and breaking process are shown in Fig. 2b, and also supported by ESI Video V1. From the enhanced thickness of the junction, the healed area can be easily recognized under cross sectional FESEM examination. However, no other imperfections or gaps were observed at the interface between the two healed strips (ESI Fig. S4). The bending stiffness of the strip calculated by employing the Lorentzen & Wettre two-point method37,38,40 before (9.6 × 10−7 N m) and after (1.07 × 10−6 N m) the healing process was found to be nearly the same, suggesting successful recovery of the mechanical properties; the details of the measurements are described in the ESI Fig. S5. Further, recovery of the mechanical properties was confirmed by tensile strength measurements of the pristine (6.3 MPa) and healed (6.6 MPa) strips, as shown in ESI Fig. S6.

image file: c8ta01885e-f2.tif
Fig. 2 Healing of Li-vermiculite membrane (a) optical microscopic images showing repairing of a hole artificially created in Li-vermiculite membrane of thickness 40 μm. (b) Photos showing multiple steps of breaking and healing of a Li-vermiculite strip. (c) Schematic illustration of the experimental set-up used for the temperature dependent healing of Li-vermiculite strips, and (d) photos of a strip healed at 140 °C. (e) Time required for healing is plotted as a function of healing temperature.

Due to its extraordinary thermal stability, vermiculite-based functional materials are usually used for high-temperature applications. Therefore, the water assisted-healing process of vermiculite strips was also examined at high temperature by placing them on a hot plate. Typically, pieces of clay strips were placed on a glass plate and heated at different temperatures (60, 80, 100, 120, 140 and 160 °C) for around 20 minutes (see the schematic of Fig. 2c). Once the thermal equilibrium was reached, 20 μL of water was dropped at the junction of the strips to facilitate the healing process. Due to the high temperature, the water droplet evaporated very rapidly and concurrently fused the broken strips. The time required for healing the Li-exchanged vermiculite as a function of healing temperature is presented in Fig. 2e. Remarkably, at 140 °C, the two vermiculite strips fused into a single piece within 30 seconds. However, heating at 160 °C or above did not heal the strips as the explosive evaporation of the water droplet caused further damage to the exposed area (images are shown in the ESI Fig. S7). In order to gain further insight into the healing time and healing efficiency, the bending stiffness of a freshly healed strip was recorded as a function of time; the result is discussed in the ESI Fig. S5c. The self-repairing process of the strip leads to the simultaneous recovery of its bending stiffness value, showing ∼87% recovery at 30 min (70 °C).

It has been demonstrated that damages in the physical structure of Li-exchanged vermiculite strips can be conveniently healed with the assistance of liquid water. However, compared to the recovery of physical structures, the healing of the materials' functionalities is considered to be a matter of greater importance. Freestanding vermiculite membranes demonstrate excellent nanofluidic ionic conductivity. In order to verify the effect of healing process on its nanofluidic conductance, the ionic conductivities of the healed strips were compared with that of pristine membranes. Nanofluidic devices were fabricated by embedding the strips (both healed and pristine) inside a freshly prepared polydimethylsiloxane (PDMS) elastomer.25,26 The current–voltage (IV) curves of pristine and healed devices with 10−2 M KCl solution are compared in Fig. 3b. The similarity in the curves convincingly reveals the recovery of nanofluidic functionality of the broken membranes by the water-assisted healing process. In the interest of further confirmation, variation of the ionic conductivities of two healed devices was compared with that of two pristine devices as a function of KCl concentration (see Fig. 3c). All four devices show identical nanofluidic characteristic of surface-charged-governed ionic conductivity with two distinct regimes, confirming the complete healing of the nanofluidic properties. Furthermore, in the high concentration regime (10−2 to 1 M), the conductivities vary linearly with varying concentrations, while in the low concentration regime (10−6 to 10−2 M), the conductivities did not change even for increase in orders of magnitude in the electrolyte concentrations. Similarly, the conductivities of pristine and healed membranes were also studied by soaking the devices in HCl solutions. The characteristics of proton conductivity, as shown in the ESI Fig. S8, do not show any noticeable difference between the two (pristine and healed strip) devices. This self-repairing quality of Li-vermiculite is also utilized to connect them with lamellar membranes of other 2D materials. An edge of a Li-vermiculite strip was placed on the top of a graphene oxide (GO) strip and a droplet of water was added to the junction. As the evaporation of water under ambient condition was completed, the vermiculite and GO strip were found to be fused into a single flexible strip (see the digital photo in Fig. 3d). The fusion of the strips is attributed to swelling and intermixing of vermiculite and GO flakes at the interface upon addition of water, and such behaviour of GO is already reported in the literature.11,12,37 Along with their physical structures, the nanofluidic conductivities of the two strips also fused in a coherent manner. As shown in Fig. 3e, nanofluidic conductivity of the GO-vermiculite strip is compared with that of pure GO and pure vermiculite membranes. ESI Fig. S9 shows a photo of the vermiculite strip fused with a strip of vanadium pentoxide lamellar membrane. The fusion of vermiculite and vanadium pentoxide strips is also attributed to a similar swelling intermixing of the components at the interface as flakes of both the components are redispersible in water.

image file: c8ta01885e-f3.tif
Fig. 3 Effect of healing on nanofluidic transport (a) schematic representation and (b) IV curves of nanofluidic devices prepared with pristine and healed Li-vermiculite strips. Ionic conductivity data as a function of salt concentration for nanofluidic devices of (c) pristine and healed strips of Li-vermiculite, and (e) Li-vermiculite fused with a GO strip. (d) Photos showing fusion of vermiculite and GO strips. (f) Humidity-dependent conductivity of the air-dried Li-vermiculite membrane. (g) Schematic representation, and (h) ionic conductivity of air-dried Li-vermiculite strips in multiple cracked and healed stages, at RH = 88%.

Even though vermiculite sheets are electrically insulators, air-dried strips of vermiculite membrane display humidity dependent conductivity (see Fig. 3f), which is attributed to the extraordinary affinity of vermiculite towards water molecules. At high atmospheric humidity levels (above 80%), a fraction of the vermiculite nanofluidic-network is filled with an interconnecting layer of water molecules, facilitating continuous flow of charged species. As the atmospheric humidity level decreases, some of the water molecules escape from nanochannels, breaking the continuity of water channels and hence, the conductivity decreases. Interestingly, this humidity dependent conductivity of the Li-exchanged vermiculite strip can also be healed with the help of water droplets. In a typical experiment, the ends of a rectangular strip of vermiculite membrane were fixed to copper wires connected to a sourcemeter instrument by employing conductive silver paste. After measuring the conductivity, the strip was cut in the middle with a sharp blade, physically disrupting the flow through the nanofluidic channels. The broken strips were placed together and a droplet of water was applied. As the water molecules evaporated the nanochannels of the strip cured itself and the conductivity value was recovered. This process was further repeated for multiple cutting and healing steps without any change in the conductivity values (shown in Fig. 3h). The schematic representation of the consecutive cutting and self-repairing processes is shown in the Fig. 3g.

Charge-balancing cations are an integral part of the clay structure and are also a major contributor to the overall properties of clay-based materials.41 Therefore, healing properties of the vermiculite strips intercalated with different charge-balancing cations have been examined. In order to exchange Li+ with other cations (K+, Na+, Mg2+, Ba2+ and Ca2+) a Li-vermiculite sample was independently stirred with 2 M aqueous solutions of KCl, NaCl and 0.2 M aqueous solutions of MgCl2, BaCl2 and CaCl2 for 12 hours, followed by washing with water through centrifugation. Vermiculite with H+ ions was prepared by stirring vermiculite crystals with 20% aqueous solution of HCl. Membranes of the respective ion-exchanged vermiculite layers were prepared as per the details given in the experimental section. The healing process was also repeated to connect broken strips of membranes with different intercalating ions. Surprisingly, except Li-exchanged vermiculite, none of the clay strips showed water-assisted healing capabilities (see the ESI Fig. S10). The Ba-vermiculite membrane, which did not show healing ability (Fig. 4a), was re-exchanged with Li+ ions and remarkably, the healing property of the vermiculite membrane was reinstated (Fig. 4b). Valuable insights on the mechanism of the water-assisted healing process of Li-exchanged vermiculite membranes were obtained through in situ monitoring of the healing area by recording XRD patterns at regular intervals of time (see Fig. 4c). Immediately after addition of a water droplet, the intensity of the (002) reflection, which is attributed to the interlayer spacing of the clay layers, shifted to a lower 2θ value from 6.54° to 5.35° (interlayer spacing (d) from 1.35 nm to 1.65 nm, respectively), along with a simultaneous decrease in the intensity of the X-ray reflection. This change suggests the swelling of the clay layers with a partial loss in the lamellar arrangement. With time, the water molecules started evaporating and concentration of the clay layers increased, leading to reassembling of the flakes into a lamellar arrangement. As a result, the intensity of the X-ray reflection at 2θ = 5.35° was enhanced. After around 2 hours, water molecules started escaping from the interlayer spaces of the clay layers, creating a system with a variety of interlayer spacing, which was revealed by the appearance of a very broad X-ray reflection extending from 2θ values from 5.4° to 6.6° in the respective XRD patterns. Eventually, after ∼240 minutes alongside the recovery of the damaged area, a reflection pattern similar to that of the pristine clay strip was recovered. A similar experiment was also repeated with vermiculite strips with other cations (K+, Na+, H+, Ba2+, Mg2+, etc.), which could not be healed with the help of water droplets. All of these strips only showed marginal shifts in the position of their (002) reflections when soaked in water, suggesting that water assisted swelling is an essential criterion for the damage-healing quality of clay strips. The comparison of the shift in the position of the (002) reflection is shown in Table 1 (ESI). Cation specific swelling of vermiculite clay is already known in the clay literature; for example, a systematic study on hydration energy of intercalating cations and expansion of vermiculite and montmorillonite clay was reported by J. A. Kittrick in the year 1969.42 The swelling nature of clay minerals is determined by the equilibrium between expansive forces such as hydration of mineral surfaces and interlayer cations and contracting forces such as ion-clay attractions. The hydration energy of vermiculite surfaces is the same for all of the vermiculite samples with different cations. Therefore, in this system, the competition is between the hydration of interlayer cations and the ion-clay attraction. Among the monovalent cations tested herein, Li+ exhibits the highest hydration energy and hydration radius, which is probably sufficient to overcome the ion-clay attraction and hence, the membrane swells in water and heals by itself. The hydration energy of other monovalent cations is not enough for the swelling of clay layers. Polyvalent cations such as Ba2+, Mg2+, and Ca2+ display hydration energies higher than that of lithium, but the higher positive charges also enhance the attraction between the interlayer cations and clay surfaces. As a net result, samples with polyvalent cations do not swell in water, which is a prerequisite for the water-assisted healing process.

image file: c8ta01885e-f4.tif
Fig. 4 Mechanism of healing photos showing multiple stages of the healing process with strips of vermiculite membranes intercalated with (a) Ba2+ ions and (b) Ba2+ replaced with Li+ ions. (c) XRD patterns of vermiculite strips at different stages of the healing process after addition of a water droplet (20 μL). (d) XRD patterns of vermiculite membranes charge balanced with different cations before and after being shocked in water.

The interlayer cation-dependent swelling of vermiculite membranes was also utilized to prepare responsive bilayer membranes by successively vacuum-filtering aqueous dispersions of exfoliated vermiculite layers exchanged with two separate charge balancing cations. Fig. 5a shows the digital image of a freestanding bilayer membrane, with one side composed of Li-exchanged vermiculite flakes and the other side made up of proton exchanged vermiculite flakes. FESEM images of the H-exchanged and Li-exchanged sides of the membrane are shown in Fig. 5b and c, respectively. As both sides of the bilayer membrane interact differently to the change in the external environment, rectangular strips (dimension, 25 mm × 2 mm × 0.012 mm) of bilayer vermiculite act as a highly sensitive shape-morphing material. Fig. 5d demonstrates the morphing of a bilayer strip upon exposure to acetone vapour. The strip sensed the presence of acetone within 0.26 s and responded by bending towards the Li+-exchanged side with a speed of 85° s−1. Once the beaker with acetone vapours was removed, the strip regained its original configuration with a speed of 135° s−1. The strip was also found to be responsive to vapours of other solvents with specific bending speeds. Fig. 5d and e represent the bending and recovery speeds of the bilayer strip upon exposure to vapours of methanol, ethanol, THF (tetrahydrofuran) and 2-propanol. With molecules of similar chemical properties the bending speed was found to be proportional to the vapour pressure of the solvents. For example, with methanol (vp = 13.0 Pa), ethanol (vp = 5.9 kPa) and 2-propanol (vp = 4.3 kPa) the bending speeds were found to be 58° s−1, 46° s−1 and 12° s−1, respectively (see Fig. 5e). Similarly, the reaction times of the strips were also found to be proportional to the boiling points of the solvents used (Fig. 5f). Shape morphing of the bilayer strips originates from the unequal alternation of the mechanical characteristics of the constituent layers. Solvent vapour-dependent bending stiffness of the H-exchanged and Li-exchanged vermiculite strip are shown in ESI Fig. S12b and c, respectively. While exposure to acetone reduces the stiffness value of the H-exchanged strips by 15%, that of Li-exchanged strip was reduced by just 5%.

image file: c8ta01885e-f5.tif
Fig. 5 Responsiveness of vermiculite–vermiculite bilayer membrane (a) digital image, (b) and (c) FESEM images of a bilayer membrane composed of Li-vermiculite and H-vermiculite. (d) Bending movement of a bilayer strip upon exposure to acetone vapours. Bar diagrams comparing (e) bending speeds, (f) recovery speeds, and (g) reaction times of vermiculite–vermiculite bilayer strips (h) digital images showing healing of a bilayer strip.

Remarkably, the responsive properties of the bilayer strip can also be recovered by applying a tiny water droplet. ESI Fig. S13 shows snapshots of vapour-assisted bending of Li and H-vermiculite bilayer strips cut into two pieces, followed by healing into a single piece. As shown in Fig. 5e–g, the bending speeds, recovery speeds and reaction times of the healed strip towards different solvents are compared with that of a pristine strip. It is truly remarkable that the bending speeds, recovery speeds, and the reaction times of the healed strip are almost same as the pristine membrane.


In conclusion, we have demonstrated that the interaction between exfoliated layers of vermiculite clay can be tuned by interchanging its charge balancing cations to prepare self-repairable nanofluidic systems. The autonomous healing ability of nanofluidic membranes would not only help to attain its true potential in commercial applications, but also create a new platform to mimic the activity of biological self-healing nanochannels. Moreover, such healing ability opens up the potential of tuning the interactions between layers of other 2D materials to prepare self-healing nanochannels of desired functionality. The ultrafast healing of Li-vermiculite membranes at high temperature would help improve its durability in harsh environmental conditions. The ability of Li-vermiculite to fuse with membranes of other 2D materials could be exploited for preparing complex nanofluidic circuits such as ionic diodes, logical devices or devices for gated ionic/molecular transport. This study also demonstrates the potential for preparing highly responsive and self-repairing smart materials by simply assembling layers of the same clay, which are charge-balanced with appropriate choices of cations.

Experimental section


Vermiculite clay was purchased from Sigma Aldrich and used as received. Hydrochloric acid, barium chloride, calcium chloride, lithium chloride, potassium chloride, sodium chloride, magnesium chloride and ethanol were purchased from Merck and used as received.

Exfoliation of vermiculite clay

Thermally expanded vermiculite crystals were exfoliated through a two-step ion exchange method. For the first step, 50 mg of clay sample was refluxed for 24 hours in a saturated solution of NaCl, followed by repeated washing with deionized (DI) water and ethanol. The Na-exchanged samples were then refluxed in 100 mL of 2 M LiCl solution for 24 hours. The as-obtained Li-exchanged vermiculite clays were repetitively washed with DI water and ethanol and used for membrane preparation.

Vermiculite with different charge-balancing cations

Vermiculite samples with K+, Na+, Ba2+, Ca2+ and Mg2+ ions were prepared independently by stirring 50 mg of Li-exchanged clays in 50 mL aqueous solutions of KCl (2 M), NaCl (2 M), BaCl2 (0.2 M), CaCl2 (0.2 M) and MgCl2 (0.2 M), respectively, for 12 hours at room temperature, followed by repetitive washing with DI water and ethanol, in sequence. The H-exchanged clay sample was prepared by stirring vermiculite clay with 20% HCl solution for 18 hours and then, it was washed with DI water until the excess acid was removed. To prepare Li-exchanged vermiculite from Ba-exchanged vermiculite, 50 mg of Ba-exchanged clay was stirred in 50 mL of 2 M aqueous solution of LiCl for 12 hours at room temperature, followed by repetitive washing with DI water and ethanol, in sequence.

Membranes were prepared by vacuum filtering 40 mL of the aqueous dispersion (1 mg mL−1) through commercial cellulose nitrite filtration membrane (47 mm in diameter) with pore diameter of 3 μm.

Preparation of vermiculite bilayer membrane

At first, 10 mL aqueous dispersions of Li-exchanged (1 mg mL−1) vermiculite flakes were vacuum filtered through a cellulose nitrite membrane (47 mm in diameter). Once the membrane of Li-exchanged vermiculite dried, 10 mL of H-exchanged vermiculite dispersion (1 mg mL−1) was filtered through it and allowed to dry at room temperature until it separated from the cellulose nitrite membrane.


The morphology of the exfoliated vermiculite layers and membranes were characterized by field emission scanning electron microscopy (FESEM) (Zeiss, Model: Sigma), atomic force microscope (AFM) (Agilent, Model 5500 series) and optical microscope (Olympus BX51). X-ray diffraction studies were carried out by using a Bruker D-205505 instrument with Cu-Kα radiation (λ = 1.5406 Å). Bending movement of the bilayer membranes were recorded by a Nikon D5200 camera. Conductivity measurements were carried out with a Keithley sourcemeter (Model: 2450).

Conflicts of interest

There are no conflicts to declare.


K. R. acknowledges financial support of Ramanujan Research Grant (SB/S2/RJN-141/2014) of the Science and Engineering Research Board (SERB), India. All the authors thank CIF-IIT Guwahati, DST-FIST-Chemistry and Start-up Grant IIT Guwahati for the help with sample characterizations. R. G., is grateful to IITG for PhD fellowships.


  1. B. C. K. Tee, C. Wang, R. Allen and Z. Bao, Nat. Nanotechnol., 2012, 7, 825 CrossRef PubMed.
  2. H. Wang, B. Zhu, W. Jiang, Y. Yang, W. R. Leow, H. Wang and X. Chen, Adv. Mater., 2014, 26, 3638 CrossRef PubMed.
  3. Y. Huang, Y. Huang, M. Zhu, W. Meng, Z. Pei, C. Liu, H. Hu and C. Zhi, ACS Nano, 2015, 9, 6242 CrossRef PubMed.
  4. J. H. Xu, S. Ye, C. D. Ding, L. H. Tan and J. J. Fu, J. Mater. Chem. A, 2018, 6, 5887 Search PubMed.
  5. B. Grabowski and C. C. Tasan, in Self-healing Materials, Springer International Publishing, Cham, 2016, p. 387 Search PubMed.
  6. S. Hautakangas, H. Schut, N. H. van Dijk, P. E. J. Rivera Díaz del Castillo and S. van der Zwaag, Scr. Mater., 2008, 58, 719 CrossRef.
  7. M. Mitomo, T. Nishimura and M. Tsutsumi, J. Mater. Sci. Lett., 1996, 15, 1976 CrossRef.
  8. K. Ando, M.-C. Chu, K. Tsuji, T. Hirasawa, Y. Kobayashi and S. Sato, J. Eur. Ceram. Soc., 2002, 22, 1313 CrossRef.
  9. R. Sugiyama, K. Yamane, W. Nakao, K. Takahashi and K. Ando, J. Intell. Mater. Syst. Struct., 2008, 19, 411 CrossRef.
  10. S. Itoh, S. Kodama, M. Kobayashi, S. Hara, H. Wada, K. Kuroda and A. Shimojima, ACS Nano, 2017, 11, 10289 CrossRef PubMed.
  11. H. Cheng, Y. Huang, Q. Cheng, G. Shi, L. Jiang and L. Qu, Adv. Funct. Mater., 2017, 27, 1703096 CrossRef.
  12. C. Luo, C. N. Yeh, J. M. L. Baltazar, C. L. Tsai and J. Huang, Adv. Mater., 2018, 1706229 CrossRef PubMed.
  13. M. Burnworth, L. Tang, J. R. Kumpfer, A. J. Duncan, F. L. Beyer, G. L. Fiore, S. J. Rowan and C. Weder, Nature, 2011, 472, 334 CrossRef PubMed.
  14. Y. Yang and M. W. Urban, Chem. Soc. Rev., 2013, 42, 7446 RSC.
  15. Y.-K. Song, Y.-H. Jo, Y.-J. Lim, S.-Y. Cho, H.-C. Yu, B.-C. Ryu, S.-I. Lee and C.-M. Chung, ACS Appl. Mater. Interfaces, 2013, 5, 1378 Search PubMed.
  16. U. Haldar, K. Bauri, R. Li, R. Faust and P. De, ACS Appl. Mater. Interfaces, 2015, 7, 8779 Search PubMed.
  17. C.-H. Li, C. Wang, C. Keplinger, J.-L. Zuo, L. Jin, Y. Sun, P. Zheng, Y. Cao, F. Lissel, C. Linder, X.-Z. You and Z. Bao, Nat. Chem., 2016, 8, 618 CrossRef PubMed.
  18. Z. Wang, L. Heng and L. Jiang, J. Mater. Chem. A, 2018, 6, 3414 Search PubMed.
  19. A. Das, J. Deka, K. Raidongia and U. Manna, Chem. Mater., 2017, 29, 8720 CrossRef.
  20. K. Golovin, M. Boban, J. M. Mabry and A. Tuteja, ACS Appl. Mater. Interfaces, 2017, 9, 11212 Search PubMed.
  21. Y. Li, Z. Zhang, M. Wang, X. Men and Q. Xue, J. Mater. Chem. A, 2017, 5, 20277 Search PubMed.
  22. A. S. Hamdy, I. Doench and H. Möhwald, Prog. Org. Coat., 2011, 72, 387 CrossRef.
  23. K. Alipour and F. Nasirpouri, RSC Adv., 2017, 7, 51879 RSC.
  24. J. Tang, L. Wan, Y. Zhou, H. Pan and F. Huang, J. Mater. Chem. A, 2017, 5, 21169 Search PubMed.
  25. K. Raidongia and J. Huang, J. Am. Chem. Soc., 2012, 134, 16528 CrossRef PubMed.
  26. J.-J. Shao, K. Raidongia, A. R. Koltonow and J. Huang, Nat. Commun., 2015, 6, 7602 CrossRef PubMed.
  27. M. Zhang, X. Hou, J. Wang, Y. Tian, X. Fan, J. Zhai and L. Jiang, Adv. Mater., 2012, 24, 2424 CrossRef PubMed.
  28. J. M. Perry, K. Zhou, Z. D. Harms and S. C. Jacobson, ACS Nano, 2010, 4, 3897 CrossRef PubMed.
  29. R. Fan, S. Huh, R. Yan, J. Arnold and P. Yang, Nat. Mater., 2008, 7, 303 CrossRef PubMed.
  30. J. Fu, R. B. Schoch, A. L. Stevens, S. R. Tannenbaum and J. Han, Nat. Nanotechnol., 2007, 2, 121 CrossRef PubMed.
  31. L. Ding, Y. Wei, L. Li, T. Zhang, H. Wang, J. Xue, L.-X. Ding, S. Wang, J. Caro and Y. Gogotsi, Nat. Commun., 2018, 9, 155 CrossRef PubMed.
  32. A. Siria, P. Poncharal, A.-L. Biance, R. Fulcrand, X. Blase, S. T. Purcell and L. Bocquet, Nature, 2013, 494, 455 CrossRef PubMed.
  33. W. Sparreboom, A. van den Berg and J. C. T. Eijkel, Nat. Nanotechnol., 2009, 4, 713 CrossRef PubMed.
  34. S. K. Min, W. Y. Kim, Y. Cho and K. S. Kim, Nat. Nanotechnol., 2011, 6, 162 CrossRef PubMed.
  35. E. T. Lam, A. Hastie, C. Lin, D. Ehrlich, S. K. Das, M. D. Austin, P. Deshpande, H. Cao, N. Nagarajan, M. Xiao and P.-Y. Kwok, Nat. Biotechnol., 2012, 30, 771 CrossRef PubMed.
  36. S. Liu, Q. Pu, L. Gao, C. Korzeniewski and C. Matzke, Nano Lett., 2005, 5, 1389–1393 CrossRef PubMed.
  37. R. K. Gogoi, K. Saha, J. Deka, D. Brahma and K. Raidongia, J. Mater. Chem. A, 2017, 5, 3523 Search PubMed.
  38. R. K. Gogoi and K. Raidongia, Adv. Mater., 2017, 29, 1701164 CrossRef PubMed.
  39. C. Marcos, Y. C. Arango and I. Rodriguez, Appl. Clay Sci., 2009, 42, 368 CrossRef.
  40. R. E. Mark, C. C. Habeger, J. Borch and M. B. Lyen, Handbook of Physical Testing of Paper, CRC Press, New York, 2001, pp. 237–239 Search PubMed.
  41. V. P. Evangelou, Environmental Soil and Water Chemistry: Principles and Applications, John Wiley & Sons, New York, 1998, pp. 103–126 Search PubMed.
  42. J. A. Kittrick, Soil Sci. Soc. Am. Proc., 1969, 33, 217 CrossRef.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c8ta01885e

This journal is © The Royal Society of Chemistry 2018